Asynchronous motor non-inductive limp control method and device based on model reference adaptive system

Abstract

This application relates to the field of motor control technology, and provides a sensorless limp-off control method and apparatus for an asynchronous motor based on a model reference adaptive system. It includes constructing a rotor flux linkage voltage model as a reference model based on stator voltage and stator current in a two-phase stationary coordinate system, and constructing a rotor flux linkage current model as an adjustable model based on stator current and speed parameters. Based on the error between the reference flux linkage component output by the reference model and the estimated flux linkage component output by the adjustable model, an adaptive law is used to obtain an estimated motor speed. In response to a speed sensor fault detection signal, the feedback speed of the speed closed-loop control is switched from the sensor feedback speed to the estimated motor speed, causing the motor to enter a limp-off operation mode. This allows for a smooth switch to a sensorless observation mode when the speed sensor fails, ensuring the motor continues to operate under load, effectively improving the fault tolerance and reliability of the drive system.

Description

Technical Field

[0001] This application relates to the field of motor control technology, specifically to a sensorless limp-off control method and device for asynchronous motors based on a model reference adaptive system. Background Technology

[0002] Three-phase asynchronous motors, with their simple structure, low cost, and reliable operation, have been widely used in industrial drives, electric vehicles, and home appliances. In mainstream vector control and direct torque control schemes, accurate speed information is a prerequisite for high-performance control. Traditionally, encoders or resolvers are installed on the motor shaft to measure speed in real time. However, the introduction of speed sensors inevitably increases system cost and installation complexity. Furthermore, the sensors themselves are prone to failure under harsh conditions such as high temperature, vibration, and dust. If the sensor fails, the entire drive system will malfunction, potentially leading to safety accidents.

[0003] To reduce reliance on physical sensors, sensorless control technology has emerged. This type of technology indirectly estimates speed and flux linkage information using easily obtainable electrical quantities from the motor itself, thus replacing the function of physical sensors to some extent. Currently, common speed estimation methods include high-frequency signal injection, sliding mode observer, extended Kalman filter, and model reference adaptive method. High-frequency signal injection requires additional high-frequency signal injection, introducing noise and losses; while the sliding mode observer is robust to parameter changes, its inherent chattering phenomenon affects steady-state accuracy; the extended Kalman filter has excellent dynamic performance, but its computational complexity is too high, making it difficult to run in real-time in low-cost embedded controllers. The model reference adaptive method uses the motor's mathematical model to construct a reference model and an adjustable model, and achieves speed identification through error driving. Its principle is clear and the computational load is moderate, but the pure integration stage involved in traditional implementations suffers from initial value sensitivity and DC bias problems, which are particularly prominent in the low-speed region, leading to a decrease in flux linkage and speed estimation accuracy.

[0004] Furthermore, most existing sensorless speed control solutions operate in an independent control mode, lacking a mechanism for dynamic switching between sensor-enabled and sensorless operation. When a speed sensor suddenly fails during normal drive system operation, if the feedback signal source cannot be smoothly switched from the sensor to the software estimate within a very short time, the motor speed and torque will fluctuate drastically or even become uncontrollable, failing to meet the limp-home functionality requirements of safety-critical applications such as electric vehicles. Therefore, providing a speed estimation scheme with better integral performance and achieving seamless switching from sensor-enabled to sensorless mode under fault conditions is a pressing issue that needs to be addressed to improve the reliability and fault tolerance of asynchronous motor drive systems. Summary of the Invention

[0005] In view of this, embodiments of this application provide a sensorless limp-off control method and apparatus for asynchronous motors based on a model reference adaptive system, which can smoothly switch to a sensorless observation mode when the speed sensor fails, ensuring that the motor continues to run under load, and effectively improving the fault tolerance and reliability of the drive system.

[0006] The first aspect of this application provides a sensorless limp-off control method for an asynchronous motor based on a model reference adaptive system, comprising: In a two-phase stationary coordinate system, a rotor flux linkage voltage model is constructed based on stator voltage and stator current as a reference model, and a rotor flux linkage current model is constructed based on stator current and speed parameters as an adjustable model. Based on the error between the reference flux component output by the reference model and the estimated flux component output by the adjustable model, the estimated motor speed is obtained through an adaptive law. In response to a speed sensor fault detection signal, the feedback speed of the speed closed-loop control is switched from the sensor feedback speed to the estimated motor speed, causing the motor to enter a limp-running mode.

[0007] A second aspect of this application provides a sensorless limp-off control device for an asynchronous motor based on a model reference adaptive system, comprising: The model building module is configured to build a rotor flux voltage model based on stator voltage and stator current as a reference model in a two-phase stationary coordinate system, and to build a rotor flux current model based on stator current and speed parameters as an adjustable model. The speed estimation module is configured to obtain the estimated motor speed value through an adaptive law based on the error between the reference flux component output by the reference model and the estimated flux component output by the adjustable model. The limp switching module is configured to switch the feedback speed of the speed closed-loop control from the sensor feedback speed to the estimated motor speed value in response to the speed sensor fault detection signal, so that the motor enters the limp operation mode.

[0008] The first aspect of this application provides a sensorless limp-off control method for asynchronous motors based on a model reference adaptive system. This method constructs a voltage model without unknown speed as a reference model and a current model with speed parameters as an adjustable model in a two-phase stationary coordinate system. The error between the flux linkage components output by the two models is used to obtain an estimated value of the motor speed through an adaptive law. When the speed sensor fails, the feedback speed is smoothly switched to the estimated value, causing the motor to enter a limp-off operation mode. This achieves seamless switching between sensor-equipped and sensorless modes, effectively solving the system downtime problem caused by a single point failure of the speed sensor in the drive system, and greatly improving the fault tolerance and reliability of the drive system.

[0009] It is understandable that the beneficial effects of the second aspect mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 This is a schematic flowchart of an embodiment of the sensorless limp-off control method for an asynchronous motor based on a model reference adaptive system provided in this application; Figure 2 This is a system principle block diagram provided in one embodiment of this application; Figure 3 This is a schematic diagram of a saturated feedback integrator structure provided in an embodiment of this application; Figure 4 This is a schematic diagram of the structure of an asynchronous motor sensorless limp-off control device based on a model reference adaptive system provided in an embodiment of this application. Detailed Implementation

[0012] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0013] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0014] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0015] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0016] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0017] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0018] The sensorless limp-off control method provided in this invention is applicable to three-phase asynchronous motor drive systems employing vector control strategies. Typical applications include electric vehicle drives, industrial frequency converters, and rail traction—fields with high requirements for system reliability and fault tolerance. In these applications, the motor controller is typically equipped with a speed sensor for closed-loop control and also has the ability to sample the motor stator voltage and stator current. The core of this invention lies in: during normal operation with the sensor, a speed estimation algorithm based on a model reference adaptive system runs in parallel; when the speed sensor fails, the feedback signal source is immediately switched from the sensor to the speed estimation value, allowing the motor to continue operating in a degraded but controllable manner, i.e., the so-called "limp-off" mode.

[0019] In one embodiment, such as Figure 1 As shown, a sensorless limp-off control method for an asynchronous motor based on a model reference adaptive system includes the following steps S101 to S103: Step S101: In a two-phase stationary coordinate system, construct a rotor flux linkage voltage model based on stator voltage and stator current as a reference model, and construct a rotor flux linkage current model based on stator current and speed parameters as an adjustable model.

[0020] Step S102: Based on the error between the reference flux component output by the reference model and the estimated flux component output by the adjustable model, obtain the estimated motor speed value through an adaptive law.

[0021] Step S103: In response to the speed sensor fault detection signal, the feedback speed of the speed closed-loop control is switched from the sensor feedback speed to the estimated motor speed, so that the motor enters the limp operation mode.

[0022] In application, steps S101 to S103 constitute the overall control flow of this invention. The two-phase stationary coordinate system refers to the system commonly used in motor analysis. A coordinate system is used, which is fixed with the stator as the reference frame. The three-phase stator voltage and current are converted into two-phase orthogonal components through Clarke transformation, thereby simplifying the expression of the mathematical model of the motor.

[0023] In step S101, the reference model refers to the rotor flux linkage model derived based on the stator voltage equation. Since the input of this model only includes the stator voltage and stator current—two directly measurable or reconstructable electrical quantities—and does not depend on the speed parameter to be identified, its output is considered an accurate reference. The adjustable model refers to the rotor flux linkage model constructed using the stator current and speed parameters as input. Since the speed parameter is an unknown quantity that needs to be estimated, the output of this model is an estimated value of the rotor flux linkage, which will adjust as the estimated speed value changes. Although the voltage model and the current model are derived through different paths, their outputs have the same physical meaning. The rotor flux component on the shaft.

[0024] In step S102, the difference between the reference flux linkage component output by the reference model and the estimated flux linkage component output by the adjustable model essentially reflects the deviation between the estimated rotational speed and the actual rotational speed. The adaptive law refers to an algorithm that dynamically adjusts the estimated rotational speed based on this flux linkage error signal, aiming to make the outputs of the two models converge, at which point the estimated rotational speed follows the actual rotational speed in real time. Figure 2 As shown, the reference model receives the stator voltage and stator current, and outputs the reference flux component. , The adjustable model receives stator current and speed estimates. Output estimated flux linkage components , The flux linkage generalized error module calculates the error signal based on the reference flux linkage component and the estimated flux linkage component. The error signal is processed by a PI controller to output an updated speed estimate. This feedback is then fed back to the adjustable model to form a closed loop, which is executed repeatedly until the error approaches zero.

[0025] In step S103, the limp-operation mode refers to the switch from a sensor-based closed-loop control mode to a sensorless closed-loop control mode with software-estimated speed feedback when the drive system detects a speed sensor fault. This allows the motor to continue operating under degraded conditions, ensuring the vehicle or equipment can safely reach the repair point. The speed sensor fault detection signal is a logic signal generated by the drive system's fault diagnosis module. When this signal is valid, the control system immediately switches the speed feedback source of the speed closed loop from the sensor's measured value to the speed estimate output by the model reference adaptive system.

[0026] The technical solution of this embodiment obtains the speed estimation value by establishing two parallel flux linkage models and using an adaptive law to drive the outputs of the two to be consistent, and achieves smooth switching of the feedback source when the sensor fails. This fundamentally solves the problem of the entire drive system failing due to a single point failure of the speed sensor, and significantly improves the fault tolerance and functional safety level of the system.

[0027] In one embodiment, the voltage model is expressed as: in, , The rotor flux linkage component on the αβ axis, , The stator voltage on the αβ axis, , The stator current is on the αβ axis. For stator resistance, For stator inductance, For rotor inductance, For mutual inductance between stator and rotor, This is the leakage inductance coefficient.

[0028] In the above expression, This means integrating the stator back electromotive force to obtain the stator flux linkage, and then subtracting the leakage flux component. And multiply by a coefficient This allows the stator flux linkage to be converted into the rotor flux linkage. Leakage inductance coefficient This reflects the leakage flux ratio caused by incomplete mutual inductive coupling between the stator and rotor inductances. Since all input quantities of this model—stator voltage and stator current—are directly obtainable electrical quantities and do not depend on motor speed parameters, it is suitable as a reference model.

[0029] In applications, the above voltage model is based on an asynchronous motor. The stator voltage equation in the coordinate system is derived from this starting point. Among them, the stator voltage... , This can be obtained through stator voltage reconfiguration. Specifically, it utilizes the DC bus voltage. The duty cycle of each phase arm within a switching cycle , , The average terminal voltage of each phase can be calculated. , , Then, from the neutral point voltage The phase voltages of each phase are obtained; finally, the Clarke transformation is used to obtain the final value. Stator voltage components on the shaft , Stator current , The value is obtained by directly measuring the current in the motor controller's current sampling circuit and then performing a Clarke transformation.

[0030] Specifically, the stator voltage is obtained through the stator voltage reconfiguration method. , The process is as follows: The stator voltage in the voltage model expression can be obtained using the stator voltage reconstruction method. .in This is the DC bus voltage. Terminal voltage, It represents the duty cycle of the circuit during one switching cycle.

[0031] The average terminal voltage during one switching cycle can be derived as follows: The neutral point voltage is: From the expressions for the average terminal voltage and the neutral point voltage during a switching cycle, the average phase voltage within one switching cycle can be obtained as follows: After obtaining the three-phase voltages, the stator voltage can be calculated using the Clarke transform. .

[0032] This embodiment provides a rotor flux calculation method that does not rely on rotational speed information through analytical derivation based on the stator voltage equation, which can provide an accurate reference benchmark for the model reference adaptive system.

[0033] In one embodiment, the expression for the current model is: in, , The estimated flux linkage component on the αβ axis output by the adjustable model. This is an estimated rotational speed. , The rotor time constant, For rotor resistance; , The reference flux linkage component output by the reference model.

[0034] In applications, the above current model is based on an asynchronous motor. The rotor voltage equation in the coordinate system is derived from this starting point. This model essentially describes the dynamic change process of the rotor flux linkage: the first term on the right-hand side of the equation... or This represents the excitation contribution of the stator current to the rotor flux linkage; the second term... or yes shaft and The cross-coupling term between shafts reflects the mutual influence between the flux linkage components of the two shafts caused by rotor rotation. It is the transfer function form of a first-order inertial element, where For the Laplace operator, The rotor time constant represents the dynamic response speed to changes in rotor flux linkage.

[0035] It should be noted that in the expression of the adjustable model, the cross-coupling term appears... , It is the output value of the reference model (voltage model), not the estimated output of the adjustable model itself. , This means that the input to the adjustable model consists of two parts: one is the directly measurable stator current. , Secondly, the flux linkage component comes from the reference model. , and the rotational speed to be estimated The output of the adjustable model , It is an estimated value of the rotor flux linkage, which is continuously adjusted. By making it approximate the actual rotational speed, the estimated flux linkage component can be made to approach the reference flux linkage component.

[0036] This embodiment constructs a current model containing rotational speed parameters as an adjustable model, making the rotational speed estimate the only adjustable parameter in the adaptive system, thus providing a clear adjustment target for the subsequent design of the adaptive law.

[0037] In one embodiment, obtaining the estimated motor speed using an adaptive law includes: The flux linkage error, characterizing the rotational speed error, is calculated based on the reference flux linkage component and the estimated flux linkage component. The motor speed estimate is output after the flux linkage error is processed by the regulator.

[0038] In applications, the design of the adaptive law is the core of the model reference adaptive system. The flux linkage error is a scalar signal composed of a reference flux linkage component and an estimated flux linkage component. This signal contains information about the deviation between the estimated and actual rotational speeds. When the reference and estimated flux linkage components are perfectly aligned, the flux linkage error is zero, and the estimated rotational speed equals the actual rotational speed. The regulator is the control element that dynamically processes the flux linkage error and outputs the estimated rotational speed. Its function is to convert the error signal into a correction value for the estimated rotational speed, driving the system to continuously reduce the gap between the two model outputs.

[0039] Specifically, within each control cycle, the controller first calculates the output of the reference model based on the latest stator voltage and stator current. , Then take the current estimated speed value Substitute into the adjustable model to calculate and estimate the flux linkage component. , Then, the flux linkage error is calculated based on the two sets of flux linkage components; finally, the flux linkage error is sent to the regulator, which outputs an updated speed estimate. This process is repeated in each control cycle, forming a closed-loop feedback structure, which causes the speed estimate to gradually converge to the actual speed.

[0040] This embodiment transforms the speed estimation problem into an error-driven problem between the outputs of two models and uses a regulator to achieve closed-loop convergence, providing a systematic speed identification framework with clear principles and simple structure.

[0041] In one embodiment, the adaptive law is designed based on Popov's hyperstability theory, and the flux linkage error is: The regulator is a PI regulator, and the estimated motor speed is: in, , The reference flux linkage component output by the reference model. , The estimated flux linkage component on the αβ axis output by the adjustable model. For integral gain, This is the proportional gain.

[0042] In applications, Popov's hyperstability theory is a classic theoretical tool for the stability analysis of nonlinear systems and is widely used in the design of model reference adaptive systems. According to Popov's hyperstability theory, adaptive PI control is used to estimate the angular velocity, which yields: The adaptive law derived from this theory guarantees the asymptotic stability of the system under a wide range of operating conditions. That is, regardless of the initial error, as long as the system parameters satisfy the positive reality condition, the estimated speed will eventually converge to the actual speed. (Magnetic flux linkage error) The physical meaning of the cross product is the cross product of the reference flux linkage vector and the estimated flux linkage vector. When the directions of the two flux linkage vectors are consistent, the cross product is zero, indicating that the estimated speed has accurately tracked the actual speed. When there is an angular deviation between the two flux linkage vectors, the cross product is not zero. Its sign and magnitude reflect whether the estimated speed is too high or too low and the degree of deviation, respectively.

[0043] A PI controller consists of a proportional gain and an integral gain. It is responsible for a rapid response to the current error, giving the system good dynamic tracking capabilities; integral gain. It is responsible for eliminating steady-state errors and ensuring that the estimated speed can accurately track the actual speed under constant-speed conditions. This is achieved through proper tuning. and The value of can achieve a good balance between dynamic response speed and steady-state accuracy.

[0044] This embodiment designs an adaptive law based on Popov's superstability theory, which theoretically guarantees the convergence and stability of speed estimation. Combined with a PI controller, it achieves fast and error-free tracking of flux linkage error, providing reliable speed identification performance for the entire model reference adaptive system.

[0045] In one embodiment, the integral operation in the voltage model is implemented using a saturated feedback integrator, which includes: The low-pass filter branch is used to filter the input signal to eliminate the influence of the initial integral value; The saturation feedback branch is used to limit the amplitude of the output flux linkage and feed it back to the input of the low-pass filter branch to suppress DC bias.

[0046] In applications, integration operations in voltage models In practical digital implementation, there are two main problems: First, the problem of initial integration value. Since the initial value of the flux linkage is difficult to determine accurately at the moment of motor start-up, the pure integrator will keep the initial value error in the output, resulting in a constant deviation in flux linkage estimation. Second, the problem of DC bias. The DC components such as zero drift and quantization error that are unavoidable in the sampling circuit will continue to accumulate after pure integration, eventually causing the integrator output to saturate and lose its working ability.

[0047] like Figure 3 As shown, the saturated feedback integrator consists of two parts: a low-pass filter branch and a saturated feedback branch. The low-pass filter branch replaces the main channel of the pure integrator with a first-order low-pass filter. Its function is to maintain the integration characteristics of low-frequency signals while automatically attenuating the influence of the initial integration value using the finite DC gain characteristic of the low-pass filter. The saturated feedback branch forms a negative feedback loop by feeding back the output flux of the low-pass filter to its input after amplitude limiting (saturation limiting). Its function is to actively suppress the accumulation of DC bias by introducing the bias amount into the input when the output flux exhibits DC bias and deviates from the normal amplitude range, thus canceling it out.

[0048] The specific working process is as follows: Let the input electromotive force be... The output of the low-pass filter branch is The output of the saturated feedback branch is The output of the saturation limiter is The final stator flux output When there is no DC bias in the output flux, the limiter does not function, and the entire integrator maintains ideal integration characteristics; when the output flux exceeds the normal amplitude range due to bias, the limiter is activated, and the feedback signal brings the bias component back to the input for correction.

[0049] Specifically, if the electromotive force is E, then: In the formula, To output the flux linkage amplitude, as can be seen from the above, when the saturation feedback integral is limited... When the value is 0, the feedback path is broken, and the integrator is equivalent to a cutoff frequency of 0. Low-pass filter; when limiting When the flux linkage is much larger than a given value, the integrator is equivalent to a pure integrator. Therefore, the effect of the saturated feedback integrator lies between that of a pure integrator and a low-pass filter, and its control effect depends on the limiting value. and low-pass filter cutoff frequency .

[0050] if If the value is too small, the output flux linkage of the integrator will deviate significantly from the actual flux linkage in both phase and amplitude. Therefore... The value is typically taken as the rated flux linkage. When there is no DC bias at the output, the limiter does not function, and the entire integrator is equivalent to pure integration. When there is DC bias at the output, the feedback of the limiter suppresses the DC bias.

[0051] This embodiment effectively overcomes the two key problems that restrict the practical application of voltage models—sensitivity to initial integration values ​​and DC bias accumulation—by introducing a saturated feedback integrator instead of a traditional pure integrator. It improves the flux linkage estimation accuracy of the voltage model in the full speed domain, especially in the low-speed region, without significantly increasing the computational load.

[0052] In one embodiment, the transfer function of the low-pass filter branch is: The transfer function of the saturated feedback branch is ,in The cutoff frequency is used; the amplitude limit is set to the rated flux linkage amplitude of the motor; when the amplitude limit is zero, the saturated feedback integrator is equivalent to a low-pass filter; when the amplitude limit is greater than the given flux linkage, the saturated feedback integrator is equivalent to a pure integrator.

[0053] In applications, the transfer function of the low-pass filter branch This indicates that it has a cutoff frequency of A first-order low-pass filter, when the frequency of the input signal is much higher than... When the output approximates a pure integral. The result; when the input signal frequency is close to or lower than When the frequency is low, the output will be attenuated, thus avoiding the problem that the gain of a pure integrator tends to infinity in the low-frequency range.

[0054] Transfer function of saturated feedback branch Indicates the output of the saturation limiter The feedback signal is generated after this filtering stage. The function of the saturation limiter can be described by the following piecewise function: when The amplitude is within the limit value When within range, Remain unchanged; when Exceeding At that time, the output is clamped to or .

[0055] Limit value The setting directly determines the operating characteristics of the saturated feedback integrator. Setting the limiting value to the motor's rated flux linkage amplitude is based on the following considerations: Under normal operating conditions, the rotor flux linkage amplitude fluctuates around the rated value, the limiter does not function, and the integrator maintains ideal integral characteristics; when a DC bias occurs causing the output flux linkage to exceed the rated amplitude, the limiter is activated, the feedback path is opened, and the bias component is introduced into the input to cancel it out. When the limiting value is zero, the feedback path is completely disconnected, and the entire structure degenerates into a simple low-pass filter. When the limiting value is much larger than the actual given flux linkage amplitude, the limiter will not function, and the feedback path will be equivalent to a direct pass. In this case, the entire structure is equivalent to a pure integrator. Therefore, the actual performance of a saturated feedback integrator lies between that of a pure integrator and a low-pass filter, and its performance can be improved by limiting the amplitude. and cutoff frequency Both parameters can be adjusted flexibly.

[0056] This embodiment provides the specific transfer functions and limit value selection strategies for each branch of the saturated feedback integrator, and clarifies the equivalent behavior of the integrator under different extreme conditions, providing a clear parameter tuning basis and theoretical analysis framework for engineering implementation.

[0057] In one embodiment, in sensor-operated mode, the rotational speed fed back by the speed sensor is used for closed-loop speed control, while the reference model, the adjustable model, and the adaptive law operate in parallel to continuously output the estimated value of the motor speed; and a speed sensor fault detection signal is generated by monitoring the continuity and / or rationality of the speed sensor signal.

[0058] In applications, sensor-enabled operation mode refers to the default control mode of the drive system under normal operating conditions. In this mode, the speed closed-loop controller uses the actual rotational speed measured by a speed sensor such as an encoder or resolver as a feedback signal. After comparing this signal with the given rotational speed, it generates a torque current command, thereby achieving precise regulation of the motor speed. Simultaneously, the reference model, adjustable model, and adaptive law of the model reference adaptive system run continuously in parallel in the background, receiving stator voltage and stator current signals in real time and continuously outputting estimated rotational speed values. However, these estimated values ​​do not participate in the closed-loop control in sensor-enabled operation mode.

[0059] The design intent of parallel operation is that when the sensor is normal, the model reference adaptive system is already in a steady-state tracking state, and its estimated speed has converged to near the actual speed. Once a sensor failure occurs and switching is required, there is no need to wait for the estimated value to converge from zero. Reliable speed feedback can be provided immediately, thereby significantly shortening the switching transition time and reducing speed and torque fluctuations at the moment of switching.

[0060] The generation of speed sensor fault detection signals relies on continuous monitoring of sensor signal quality. Continuity detection refers to checking for abnormalities such as sudden changes, jumps, or prolonged periods without updates in the sensor signal; reasonableness detection involves comparing the speed value fed back by the sensor with the current operating conditions of the motor (such as given speed, current amplitude, etc.) to determine whether it is within a reasonable range. When continuity and / or reasonableness detection determine that the sensor signal is abnormal, a fault detection signal is generated, triggering subsequent limp-mode switching.

[0061] This embodiment provides a hot-backup speed estimation value and a reliable fault triggering mechanism for limp mode switching by running the speed estimation algorithm in parallel in sensor mode and combining the monitoring of the continuity and rationality of sensor signals, thus ensuring the timeliness and smoothness of the switching.

[0062] In one embodiment, in response to the speed sensor fault detection signal, the rotor flux angle required for vector control is switched from the integral value based on the sensor speed to the angle value calculated based on the estimated flux component of the adjustable model, and at the switching time, the feedback speed and / or rotor flux angle are numerically held or inertial filtered to ensure signal continuity.

[0063] In applications, vector control requires accurate rotor flux linkage angles to achieve directional decoupling of the stator current, that is, to decompose the stator current into an excitation component that generates magnetic flux and a torque component that generates torque. In sensor-equipped mode, the rotor flux linkage angle is typically obtained by integrating the speed measured by the sensor and superimposing the slip angle; in sensorless mode, the estimated flux linkage component output by an adjustable model is required. , The rotor flux linkage angle is calculated directly through arctangent calculation.

[0064] During switching, to avoid drastic fluctuations in motor torque and current caused by sudden changes in feedback speed and rotor flux angle, smooth transition measures are required. Numerical hold refers to locking the feedback speed and / or rotor flux angle to their values ​​at the moment of switching, gradually releasing them after the new signal source stabilizes. Inertial filtering involves applying first-order inertial filtering to the signals before and after switching, using the filter's time constant to smoothly transition the signal to the new signal source over a certain transition time, rather than abruptly changing it. In practical applications, numerical hold, inertial filtering, or a combination of both can be selected based on the system's dynamic response requirements.

[0065] This embodiment addresses another key feedback quantity in vector control besides speed, the rotor flux angle, and provides a method for switching from sensor-equipped to sensorless operation. By using numerical hold or inertial filtering, the continuity of the signal during the switching process is ensured, thereby achieving a truly seamless switching from sensor-equipped mode to sensorless limp mode and avoiding motor torque pulsation and current surges caused by signal abrupt changes.

[0066] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0067] This application also provides a sensorless limp-off control device for an asynchronous motor based on a model reference adaptive system, used to execute the steps in the knowledge distillation method embodiments described above. The knowledge distillation device can be a virtual appliance in an electronic device, run by the processor of the electronic device, or it can be the electronic device itself.

[0068] like Figure 4 As shown in the embodiment of this application, the sensorless limp-off control device 100 for an asynchronous motor based on a model reference adaptive system includes: The model building module 101 is configured to build a rotor flux voltage model based on stator voltage and stator current as a reference model in a two-phase stationary coordinate system, and to build a rotor flux current model based on stator current and speed parameters as an adjustable model. The speed estimation module 102 is configured to obtain the estimated motor speed value through an adaptive law based on the error between the reference flux component output by the reference model and the estimated flux component output by the adjustable model. The limp switching module 103 is configured to switch the feedback speed of the speed closed-loop control from the sensor feedback speed to the estimated motor speed value in response to the speed sensor fault detection signal, so that the motor enters the limp operation mode.

[0069] In application, the above-described device represents a functional modularization of the aforementioned method embodiments. The model building module is responsible for real-time calculation of the outputs of the reference model and the adjustable model, and internally includes a voltage model calculation unit and a current model calculation unit, corresponding to the reference model and the adjustable model in the aforementioned embodiments, respectively. The speed estimation module is responsible for calculating the flux linkage error based on the outputs of the two models and outputting an estimated speed value through an adaptive law; internally, it includes an error calculation unit and a PI controller. The limp-switching module is responsible for receiving fault detection signals and performing feedback source switching operations; internally, it includes a fault detection unit and a signal switching unit.

[0070] In applications, the above modules can be implemented in software within the digital signal processor or microcontroller of the motor controller, or in hardware such as a field-programmable gate array (FPGA), or a combination of software and hardware. Data interaction between modules is accomplished through the controller's internal data bus or shared memory.

[0071] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A sensorless limp-off control method for an asynchronous motor based on a model reference adaptive system, characterized in that, include: In a two-phase stationary coordinate system, a rotor flux linkage voltage model is constructed based on stator voltage and stator current as a reference model, and a rotor flux linkage current model is constructed based on stator current and speed parameters as an adjustable model. Based on the error between the reference flux component output by the reference model and the estimated flux component output by the adjustable model, the estimated motor speed is obtained through an adaptive law. In response to a speed sensor fault detection signal, the feedback speed of the speed closed-loop control is switched from the sensor feedback speed to the estimated motor speed, causing the motor to enter a limp-running mode.

2. The method according to claim 1, characterized in that, The expression for the voltage model is: in, , The rotor flux linkage component on the αβ axis, , The stator voltage on the αβ axis, , The stator current is on the αβ axis. For stator resistance, For stator inductance, For rotor inductance, For mutual inductance between stator and rotor, This is the leakage inductance coefficient.

3. The method according to claim 1, characterized in that, The expression for the current model is: in, , The estimated flux linkage component on the αβ axis output by the adjustable model. This is an estimated rotational speed. , The rotor time constant, Rotor resistance; , The reference flux linkage component output by the reference model.

4. The method according to claim 1, characterized in that, The process of obtaining the estimated motor speed using an adaptive law includes: The flux linkage error, characterizing the rotational speed error, is calculated based on the reference flux linkage component and the estimated flux linkage component. The motor speed estimate is output after the flux linkage error is processed by the regulator.

5. The method according to claim 4, characterized in that, The adaptive law is designed based on Popov's hyperstability theory, and the flux linkage error is: The regulator is a PI regulator, and the estimated motor speed is: in, , The reference flux linkage component output by the reference model. , The estimated flux linkage component on the αβ axis output by the adjustable model. For integral gain, This is the proportional gain.

6. The method according to claim 2, characterized in that, The integral operation in the voltage model is implemented using a saturated feedback integrator, which includes: The low-pass filter branch is used to filter the input signal to eliminate the influence of the initial integral value; The saturation feedback branch is used to limit the amplitude of the output flux linkage and feed it back to the input of the low-pass filter branch to suppress DC bias.

7. The method according to claim 6, characterized in that, The transfer function of the low-pass filter branch is The transfer function of the saturated feedback branch is ,in The cutoff frequency; The amplitude limit is set to the rated flux linkage amplitude of the motor; When the limiting value is zero, the saturated feedback integrator is equivalent to a low-pass filter; when the limiting value is greater than a given flux linkage, the saturated feedback integrator is equivalent to a pure integrator.

8. The method according to claim 1, characterized in that, Also includes: In sensor-operated mode, the speed closed-loop control is performed using the rotational speed feedback from the speed sensor. The reference model, the adjustable model, and the adaptive law operate in parallel to continuously output the estimated motor speed. The speed sensor fault detection signal is generated by monitoring the continuity and / or rationality of the speed sensor signal.

9. The method according to claim 1, characterized in that, In response to the speed sensor fault detection signal, the rotor flux angle required for vector control is switched from the integral value based on the sensor speed to the angle value calculated based on the estimated flux component of the adjustable model. At the switching moment, the feedback speed and / or rotor flux angle are numerically held or inertial filtered to ensure signal continuity.

10. A sensorless limp-off control device for an asynchronous motor based on a model reference adaptive system, characterized in that, include: The model building module is configured to build a rotor flux voltage model based on stator voltage and stator current as a reference model in a two-phase stationary coordinate system, and to build a rotor flux current model based on stator current and speed parameters as an adjustable model. The speed estimation module is configured to obtain the estimated motor speed value through an adaptive law based on the error between the reference flux component output by the reference model and the estimated flux component output by the adjustable model. The limp switching module is configured to switch the feedback speed of the speed closed-loop control from the sensor feedback speed to the estimated motor speed value in response to the speed sensor fault detection signal, so that the motor enters the limp operation mode.

Images