Fractional-order integral type steering motor position and speed estimation method and related device

Abstract

The application provides a fractional integral type steering motor position and speed estimation method and related equipment, which comprises the following steps: based on three-phase currents and three-phase voltages of a target steering motor, an extended back electromotive force state observer is used to obtain a static two-phase back electromotive force estimation value in an alpha-beta static coordinate system; the static two-phase back electromotive force estimation value is rotated to obtain a rotating two-phase back electromotive force estimation value, and a half-sine integral type position error signal is constructed; the position error signal is input into a fractional integral type speed loop with an adjustable fractional order to obtain a target speed estimation value; and an integral order integral operation is performed on the target speed estimation value to obtain a target position estimation value. Thus, by introducing the fractional integral type speed loop with the adjustable fractional order, the accuracy and stability of the steering motor position and speed estimation are improved.

Description

Technical Field

[0001] This application relates to the field of motor control technology, and in particular to a fractional integral type method and related equipment for estimating the position and speed of a steering motor. Background Technology

[0002] In sensorless control of steering motors, accurate estimation of rotor position and operating speed is a core prerequisite for achieving stable control. Existing methods typically employ phase-locked loops (PLLs) and sliding mode observers to extract back electromotive force and calculate the angle. However, the application scenarios of steering motors are characterized by low-speed operation, frequent commutation, and small-angle oscillation, leading to problems such as phase jumps and chattering interference in traditional methods.

[0003] To improve estimation stability, existing technologies employ integral filters to process error signals, smoothing noise and outputting speed estimates through integration. However, conventional integer-order integral estimators handle error information in a relatively simplistic way. Their fixed integral structure struggles to achieve a flexible balance between rapid dynamic response and steady-state disturbance rejection performance when facing complex operating conditions such as motor parameter fluctuations and external load disturbances. When the steering motor is in a state of frequent commutation or sudden load changes, the integer-order integral accumulates all historical errors with equal weights, failing to adjust the dependence on recent and long-term errors according to changes in operating conditions. This results in a lag in speed estimation convergence during dynamic processes, making it difficult to quickly track changes in actual rotational speed. Furthermore, the equal-weight accumulation characteristic limits its ability to suppress disturbance signals, making it prone to estimation drift in the steady-state phase. Therefore, enhancing the flexibility of error information processing while retaining the smoothing advantages of the integral structure, enabling the estimator to adaptively adjust its response characteristics according to changes in operating conditions, is crucial for further improving the performance of sensorless steering motor control. Summary of the Invention

[0004] The purpose of this application is to at least address one of the aforementioned technical deficiencies, particularly the technical deficiency in the prior art of how to enhance the flexible processing capability of error information while retaining the smoothness advantage of the integral structure, so that the estimator can adaptively adjust its response characteristics according to changes in operating conditions.

[0005] In a first aspect, this application provides a fractional integral-type method for estimating the position and speed of a steering motor, the method comprising:

[0006] Based on the three-phase current and three-phase voltage of the target steering motor, the estimated values ​​of the two stationary back electromotive forces in the αβ stationary coordinate system are obtained by using a pre-constructed back electromotive force extended state observer.

[0007] Rotate the two stationary opposite electromotive force estimates to obtain the two rotated opposite electromotive force estimates, and construct a half-sine integral type position error signal;

[0008] The position error signal is input into a fractional-order integral velocity loop with an adjustable fractional order to obtain the target velocity estimate.

[0009] The target velocity estimate is obtained by performing an integer-order integral operation on the target velocity estimate.

[0010] In one embodiment, the back EMF extended state observer is built based on the voltage balance model of the steering motor and is used to estimate the back EMF as an extended state in real time.

[0011] In one embodiment, the step of obtaining the estimated values ​​of the two stationary back electromotive forces in the αβ stationary coordinate system based on the three-phase current and three-phase voltage of the target steering motor using a pre-built back electromotive force extended state observer includes:

[0012] Clark transformations were performed on the three-phase current and three-phase voltage of the target steering motor to obtain the actual current signal and actual voltage signal in the αβ stationary coordinate system.

[0013] Based on the pre-constructed back EMF extended state observer, state estimation is performed using actual current and voltage signals to obtain the estimated current signal. Based on the deviation between the estimated current signal and the actual current signal, the estimated values ​​of the two stationary back EMFs in the αβ stationary coordinate system are calculated.

[0014] In one embodiment, the step of rotating two stationary back EMF estimates to obtain two rotated back EMF estimates and constructing a half-sine integral position error signal includes:

[0015] By performing a Park transformation on the estimated values ​​of the two opposing electromotive forces at rest, we obtain the estimated values ​​of the two opposing electromotive forces in the rotating dq coordinate system.

[0016] Calculate the estimated value of the back electromotive force amplitude based on the estimated values ​​of the two rotating back electromotive forces;

[0017] A half-sine integral type position error signal is constructed based on the specific axis components in the back EMF amplitude estimate and the back EMF estimates of the two rotating sides.

[0018] In one embodiment, the formula for the target velocity estimate is:

[0019]

[0020] in, This represents the estimated target speed. This represents the proportional parameter of the fractional integral velocity loop. The integral parameters represent the fractional integral velocity loop. Denotes the Laplace field operator, This represents the fractional integral in the corresponding time domain. This indicates the adjustable fractional order. This indicates the position error signal.

[0021] In one embodiment, the fractional integral is defined using the Caputo operator, which is obtained by performing a convolution integral on the position error signal at historical time points with a power-law weighted function.

[0022] In one embodiment, the formula for the target location estimate is:

[0023]

[0024] in, This represents the estimated location of the target. This represents the integer integral in the corresponding time domain.

[0025] Secondly, this application provides a fractional integral type steering motor position and speed estimation device, the device comprising:

[0026] The stationary two-phase back electromotive force estimation module is used to obtain the stationary two-phase back electromotive force estimates in the αβ stationary coordinate system based on the three-phase current and three-phase voltage of the target steering motor and using a pre-built back electromotive force extended state observer.

[0027] The position error signal construction module is used to rotate the two stationary opposite electromotive force estimates to obtain the two rotated opposite electromotive force estimates, and construct a half-sine integral type position error signal.

[0028] The target velocity estimation module is used to input the position error signal into a fractional-order integral velocity loop with an adjustable fractional order to obtain the target velocity estimate.

[0029] The target position estimation module is used to perform integer-order integration on the target velocity estimate to obtain the target position estimate.

[0030] Thirdly, this application provides a storage medium storing computer-readable instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of any of the fractional integral type steering motor position and speed estimation methods described in the above embodiments.

[0031] Fourthly, this application provides a computer device, including: one or more processors, and a memory;

[0032] The memory stores computer-readable instructions that, when executed by one or more processors, perform the steps of any of the fractional integral type steering motor position and speed estimation methods described in the above embodiments.

[0033] As can be seen from the above technical solutions, the embodiments of this application have the following advantages:

[0034] The fractional-order integral-type method and related equipment for estimating the position and speed of a steering motor provided in this application first obtains the estimated values ​​of the stationary two back EMFs in the αβ stationary coordinate system based on the three-phase current and three-phase voltage of the target steering motor using a pre-constructed back EMF extended state observer. Then, the estimated values ​​of the stationary two back EMFs are rotated to obtain the estimated values ​​of the rotating two back EMFs, and a half-sine integral-type position error signal is constructed. Subsequently, the position error signal is input into a fractional-order integral-type speed loop with an adjustable fractional order to obtain the estimated value of the target speed. This speed loop uses a fractional-order integral operator with an adjustable fractional order to process the input error signal, making its structure different from the fixed structure of conventional integer-order integral-type estimators. Since the weighting characteristics built into the fractional-order integral operator can be adjusted by the fractional order, when the operating conditions change, the speed loop can achieve flexible processing of error information through the fractional order, enabling the speed estimate to quickly respond to changes in the actual rotational speed during dynamic processes and effectively suppressing estimation drift caused by disturbances during steady-state processes. This avoids the convergence lag and steady-state drift problems caused by the fixed structure of conventional integer-order integral-type estimators. Finally, integer-order integration is performed on the target velocity estimate to obtain the target position estimate. Therefore, this application solves the technical problem of conventional integer-order integral estimators struggling to balance dynamic response speed and steady-state disturbance rejection performance under complex operating conditions by introducing a fractional-order integral velocity loop with adjustable fractional order, thus improving the accuracy and stability of steering motor position and velocity estimation. Attached Figure Description

[0035] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, 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.

[0036] Figure 1 A flowchart illustrating the fractional integral type steering motor position and speed estimation method provided in this application embodiment;

[0037] Figure 2 A schematic diagram of the structure of the fractional integral type steering motor position and speed estimation device provided in the embodiments of this application;

[0038] Figure 3 This is a schematic diagram of the internal structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0039] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0040] This application provides a fractional integral method for estimating the position and speed of a steering motor. The following embodiments illustrate this method using a computer device as an example. It is understood that the computer device can be any device with data processing capabilities, including but not limited to a single server, server cluster, personal laptop, desktop computer, etc. Figure 1 As shown, the method includes:

[0041] S101: Based on the three-phase current and three-phase voltage of the target steering motor, the estimated values ​​of the two stationary back electromotive forces in the αβ stationary coordinate system are obtained using a pre-built back electromotive force extended state observer.

[0042] Here, three-phase current refers to the instantaneous current flowing through the three-phase stator windings of the target steering motor, and three-phase voltage refers to the instantaneous voltage applied to the three-phase stator windings of the target steering motor. The back EMF extended state observer is a dynamic system built upon a mathematical model of the motor. Its function is to take the three-phase current and three-phase voltage as inputs and estimate the back EMF induced in the stator windings during motor operation in real time through an internal state reconstruction mechanism. The αβ stationary coordinate system is a spatially fixed two-phase orthogonal coordinate system. Coordinate transformation maps the three-phase physical quantities of the motor to this coordinate system. The estimated stationary two-phase back EMF values ​​refer to the two back EMF components along the α-axis and β-axis directions obtained in the αβ stationary coordinate system after processing by the back EMF extended state observer.

[0043] In the specific implementation process, the instantaneous values ​​of the three-phase current of the target steering motor are first acquired in real time by current sensors installed inside the steering motor controller, and the instantaneous values ​​of the three-phase voltage applied to the three-phase winding terminals of the motor are simultaneously obtained by voltage detection circuits. To ensure the synchronization and accuracy of the data, these current and voltage signals are typically converted into digital quantities by analog-to-digital converters before being input to the microcontroller's computing core. Inside the microcontroller, an extended state observer based on the electrical parameters and mathematical model of the target steering motor is pre-built. The observer is constructed based on the voltage equation of the motor in the αβ stationary coordinate system, and incorporates the back electromotive force (EMF) as an extended state variable into the observer structure. When the digitized three-phase current and three-phase voltage are input to the observer, the observer recursively calculates in real time according to the built-in algorithm, gradually converges, and outputs the estimated values ​​of the two stationary back EMFs in the αβ stationary coordinate system. This calculation process is essentially an online reconstruction of the key physical quantity of the motor's back EMF, providing the core input for subsequent rotor position and speed estimation. Because the back EMF extended state observer itself has a certain disturbance rejection capability, even under conditions where there is noise in the current sampling or voltage fluctuations, the obtained static back EMF estimates can still maintain a high signal-to-noise ratio and estimation accuracy.

[0044] In practical applications of steering motors, such as electric power steering systems, the motors often operate at low speeds, with frequent reversals and small-angle oscillations. Under these conditions, the back EMF signal amplitude is small and easily affected by interference, placing stringent demands on the accuracy and robustness of the estimation method. By directly acquiring the estimated back EMF values ​​of two stationary back EMFs through an extended state observer, the prior information of the motor model can be fully utilized to suppress measurement noise and model mismatch, thus obtaining a relatively pure back EMF waveform even under low signal-to-noise ratio conditions. This process does not rely on additional mechanical position sensors; online identification of the back EMF can be achieved using only electrical quantities, aligning with the technical approach of sensorless control. Furthermore, because the observer adopts an extended state form, estimating the back EMF as a state variable in real time is equivalent to incorporating parameter changes and unmodeled dynamics into the observation scope, enhancing the estimation method's adaptability to motor parameter perturbations and external disturbances.

[0045] It should be noted that by processing the three-phase current and three-phase voltage using an extended state observer for back EMF, real-time online estimation of the back EMF can be achieved without mechanical position sensors, directly obtaining the estimated values ​​of the stationary two-phase back EMF in the αβ stationary coordinate system. Using an extended state observer to reconstruct the back EMF as a state variable effectively suppresses the interference of current sampling noise and voltage fluctuations on the estimation results, thereby improving the signal-to-noise ratio and stability of the back EMF estimate. Extending the back EMF as a state variable for estimation also compensates for the effects of motor parameter perturbations and external disturbances to a certain extent, making the estimation process more adaptable to changes in actual operating conditions. The obtained stationary two-phase back EMF estimates have high accuracy and smoothness, providing a reliable data foundation for subsequent rotor position and speed calculations based on the back EMF.

[0046] S102: Rotate the two stationary opposite electromotive force estimates to obtain the two rotating opposite electromotive force estimates, and construct a half-sine integral type position error signal.

[0047] The rotating back EMF estimates refer to the two back EMF components along the direct and quadrature axes obtained in the synchronous rotating coordinate system after the stationary back EMF estimates in the stationary coordinate system have undergone a coordinate rotation transformation based on the real-time rotor position angle. The half-sine integral position error signal refers to the signal quantity used to characterize the rotor position estimation deviation, constructed by integrating the quadrature axis component of the rotating back EMF estimates within an integration interval corresponding to the half-wave width of the motor's electrical cycle.

[0048] In the specific implementation process, the rotor position estimate of the target steering motor is first obtained from the previous control cycle. This position estimate can be the position result obtained from the previous calculation or the value set during initialization. A rotating coordinate transformation matrix is ​​constructed using this rotor position estimate, and a Parker transformation is performed on the current stationary back EMF estimates to calculate the direct-axis back EMF and quadrature-axis back EMF estimates in the synchronous rotating coordinate system. This transformation process essentially maps the back EMF signal, which contains sinusoidal fluctuations, in the stationary coordinate system to a coordinate system that rotates synchronously with the rotor, establishing a direct correlation between the quadrature-axis back EMF component and the rotor position estimation error. Subsequently, the time length corresponding to the motor's electrical cycle is calculated based on the currently estimated rotor electrical angular velocity, and half of this time length is taken as the width of the integration window. Within this integration window, the quadrature-axis back EMF estimates are continuously accumulated or numerically integrated. When the integration window slides to the current moment, the result of the integration operation is the half-sine integral type position error signal. Considering that the steering motor has a long electrical cycle when running at low speeds, the integral window width may be too large. In this case, a preset maximum integral time limit can be set to ensure that the integral calculation can be completed within a reasonable dynamic response time.

[0049] In actual operating scenarios of steering motors, such as when the electric power steering system is turning in place or oscillating at a small angle, the back EMF signal amplitude is small and easily affected by current sampling noise and inverter nonlinearity. Through the aforementioned rotating coordinate transformation, information related to rotor position can be decoupled from the high-frequency fluctuating static back EMF and concentrated in the quadrature-axis back EMF component, creating conditions for subsequent error extraction. Constructing the position error signal using a half-sine integral method is equivalent to averaging the quadrature-axis back EMF over a half-wave timescale of one electrical cycle, effectively filtering out instantaneous fluctuations caused by measurement noise and back EMF harmonics, making the output error signal more reflective of the true position deviation trend. This integral processing method fully utilizes the waveform characteristics of the back EMF signal within the electrical cycle, exhibiting stronger anti-interference capabilities compared to a simple instantaneous value. Simultaneously, the adaptive adjustment of the integral window ensures that a reasonable integration effect is maintained even when the rotational speed changes.

[0050] It should be noted that by transforming the stationary back EMF into direct-axis and quadrature-axis components in a rotating coordinate system through coordinate rotation, the rotor position information is decoupled and extracted, providing a clear physical quantity for subsequent quantification of position deviation. Processing the quadrature-axis back EMF using a half-sine integral and averaging it over half the width of one electrical cycle significantly suppresses the disturbances caused by instantaneous noise and back EMF harmonics on the error signal, improving the smoothness and signal-to-noise ratio of the error signal. This integral construction method matches the periodic characteristics of the steering motor's back EMF, enabling the error signal to more accurately reflect the average trend of the rotor position estimation deviation. This provides a high-quality, low-fluctuation input signal for the subsequent stable estimation of the speed loop, thereby improving the estimation stability of the entire sensorless control system under low-speed, commutation conditions.

[0051] S103: Input the position error signal into a fractional-order integral velocity loop with an adjustable fractional order to obtain the target velocity estimate.

[0052] Among them, the fractional-order integral speed loop with adjustable fractional order refers to a signal processing structure based on fractional calculus theory. This structure performs fractional-order integration on the input position error signal, and the order of this integration is an adjustable parameter. The target speed estimate is the value output after processing by this fractional-order integral speed loop, which represents the estimation result of the real-time rotational speed of the target steering motor.

[0053] In the specific implementation process, a fractional-order integration algorithm module is first pre-built inside the microcontroller. The core of this module is a discretizable fractional-order integration operator, whose integration order is designed as a variable that can be adjusted in real time. The half-sine integral position error signal constructed in the previous step is fed into this fractional-order integral speed loop as input. During the fractional-order integration operation, unlike traditional integer-order integration which accumulates all historical error information with equal weight, the fractional-order integration operator assigns different weight coefficients to historical error information according to a pre-set fractional-order value. This fractional-order value can be dynamically adjusted according to the real-time operating conditions of the target steering motor. For example, during dynamic processes such as frequent commutation or sudden load changes in the steering motor, the fractional-order value can be reduced, making the fractional-order integration operator focus more on the impact of recent error information, thereby accelerating the speed loop's response speed to changes in input error and enabling the output target speed estimate to quickly follow the rapid changes in the actual rotational speed. When the steering motor is in a steady-state condition of uniform rotation or stationary position, the fractional-order value can be increased, allowing the fractional-order integral operator to assign higher weight to long-term error information. This enhances the smoothing and filtering effect on residual disturbances and noise in the input error signal, suppressing steady-state drift of the target speed estimate. The discretization of the fractional-order integral operation can be achieved using numerical algorithms based on generating functions or rational approximation. By iteratively updating the integral state variables in each control cycle, the target speed estimate for the current moment can be output in real time.

[0054] In practical applications of electric power steering systems, when the driver performs rapid steering maneuvers or the vehicle travels on bumpy roads causing sudden load changes, the motor speed needs to respond quickly. At this time, by adjusting the fractional order of the fractional integral speed loop to a smaller value, the speed estimate can quickly track the actual speed change, avoiding a decrease in control performance due to estimation lag. When the driver holds the steering wheel at a certain angle or the vehicle is traveling in a straight line, the motor enters a near-steady-state operating state. At this point, adjusting the fractional order to a larger value utilizes the strong smoothing characteristics of the fractional integral to effectively filter out minor error disturbances introduced by back EMF harmonics or current sampling noise, making the output speed estimate more stable and avoiding additional torque pulsations caused by speed estimation fluctuations. This ability to adjust the fractional order in real time according to operating conditions allows the same speed loop structure to adapt to different performance requirements across a wide operating range of the steering motor.

[0055] It should be noted that processing the input error signal using a fractional-order integrator breaks the structural limitations of traditional integer-order integrators with equal weighted accumulation, introducing adjustable weighting characteristics. Since the fractional order can be adjusted in real time according to changes in operating conditions, the speed loop can switch its response characteristics for different operating states. In dynamic processes, lowering the fractional order increases sensitivity to error changes, enabling rapid tracking of the speed estimate. In steady-state processes, increasing the fractional order enhances the ability to suppress disturbances, reducing fluctuations and drift in the estimate. This dual advantage, achieved within the same structure, avoids the inherent problem of traditional fixed-structure integrators struggling to balance dynamic response speed and steady-state disturbance rejection performance. The final target speed estimate possesses both rapid following characteristics and steady-state smoothness, providing more accurate and reliable speed feedback information for subsequent rotor position estimation and closed-loop control.

[0056] S104: Perform integer-order integration on the target velocity estimate to obtain the target position estimate.

[0057] Integer-order integration refers to the first-order integration process of continuously accumulating or summing the input signal over time, and its mathematical essence corresponds to integer-order integration in the traditional sense. The target position estimate is the output value obtained after performing integer-order integration on the target velocity estimate as the input quantity. This value represents the estimation result of the real-time angular position of the target steering motor rotor.

[0058] In the specific implementation process, the target velocity estimate obtained in the previous step through real-time calculation using a fractional-order integral velocity loop is input to the integer-order integrator. This integer-order integrator is implemented in a discretized form within the microcontroller. Its basic operation method is to multiply the current target velocity estimate by the length of the control cycle within each control cycle to obtain the angle increment for that cycle. This angle increment is then added to the target position estimate accumulated in the previous control cycle to update the current target position estimate. This process is repeated in each control cycle to achieve continuous tracking and accumulation of the target velocity estimate over time. To ensure the accuracy of the initial integral value, an initial angle value can be set based on the initial position of the steering motor during system startup or reset, or integration can start from zero and rely on subsequent back electromotive force information to quickly converge the position estimate. In practical applications, due to the discrete nature of digital computation, integer-order integration can employ various numerical integration methods such as rectangular integration or trapezoidal integration. The choice of different integration methods will affect the accuracy of angle accumulation, and appropriate configuration can be made according to the sampling frequency and accuracy requirements of the control system.

[0059] In real-world operating scenarios of electric power steering systems, such as when a vehicle is using parking assist or lane keeping assist, the rotor position of the steering motor needs to be continuously monitored to determine the steering angle. By performing integer integration on the target speed estimate, which has already undergone fractional integral speed loop smoothing, the cumulative rotation angle of the rotor can be obtained in real time. This process converts speed information into position information, providing direct feedback for the angle closed-loop control of the steering system. When the driver performs a rapid steering operation, the target speed estimate can quickly follow the speed change, and the target position estimate obtained after integer integration can also reflect the actual position change of the rotor in real time without significant lag. After the system enters steady state, because the target speed estimate itself has already undergone fractional integral smoothing, its fluctuations and drifts are small. The target position estimate accumulated after integer integration also maintains good stability, avoiding abnormal jumps or excessive cumulative errors in the position estimate due to speed estimation fluctuations. Integer integration itself has a simple structure and low computational load, which can meet the real-time calculation requirements at high control cycle frequencies without placing an excessive computational burden on the microcontroller.

[0060] It should be noted that a deterministic mathematical relationship between rotational speed and position is established through integer-order integration, giving the conversion process from speed information to position information a clear physical meaning and mathematical foundation. Integer-order integration inherently possesses smoothing and filtering characteristics, further suppressing residual high-frequency components in the input target speed estimate, resulting in a smoother and more continuous output target position estimate. Since the target speed estimate has already undergone fractional-order integral speed loop processing in the preceding steps, exhibiting good dynamic tracking performance and steady-state smoothness, the target position estimate obtained by integer-order integration inherits these advantages. It can quickly follow rotor position changes during dynamic processes and maintain stability in steady-state processes, avoiding position drift. The final target position estimate provides accurate rotor position feedback for the steering motor's vector control algorithm, a crucial prerequisite for achieving precise torque and angle control in a sensorless control system.

[0061] In the above embodiment, firstly, based on the three-phase current and three-phase voltage of the target steering motor, the estimated values ​​of the stationary two-phase back EMF in the αβ stationary coordinate system are obtained using a pre-constructed back EMF extended state observer. Further, the estimated values ​​of the stationary two-phase back EMF are rotated to obtain estimated values ​​of the rotating two-phase back EMF, and a half-sine integral position error signal is constructed. Subsequently, the position error signal is input into a fractional-order integral speed loop with an adjustable fractional order to obtain the target speed estimate. This speed loop uses a fractional-order integral operator with an adjustable fractional order to process the input error signal, making its structure different from the fixed structure of conventional integer-order integral estimators. Since the weighting characteristics built into the fractional-order integral operator can be adjusted through the fractional order, when the operating conditions change, this speed loop can achieve flexible processing of error information through the fractional order, enabling the speed estimate to quickly respond to changes in the actual rotational speed during dynamic processes and effectively suppress estimation drift caused by disturbances during steady-state processes. This avoids the convergence lag and steady-state drift problems caused by the fixed structure of conventional integer-order integral estimators. Finally, integer-order integration is performed on the target velocity estimate to obtain the target position estimate. Therefore, this application solves the technical problem of conventional integer-order integral estimators struggling to balance dynamic response speed and steady-state disturbance rejection performance under complex operating conditions by introducing a fractional-order integral velocity loop with adjustable fractional order, thus improving the accuracy and stability of steering motor position and velocity estimation.

[0062] In one embodiment, the back EMF extended state observer is built based on the voltage balance model of the steering motor and is used to estimate the back EMF as an extended state in real time.

[0063] In this embodiment, the voltage balance model of the steering motor refers to the equations describing the mathematical relationship between stator voltage, stator current, and back electromotive force (EMF) based on the motor's electrical characteristics. This model reflects the fundamental physical laws governing the changes in electromagnetic field and electrical quantities during motor operation. The back EMF extended state observer is a dynamic system built upon this voltage balance model. Its core design idea is to incorporate the back EMF, which is not originally a state variable in the model, as an additional extended state into the observer structure. By processing the input three-phase current and three-phase voltage information in real time, the extended system state is reconstructed online, thereby achieving continuous estimation of the back EMF.

[0064] The observer constructed based on the voltage balance model of the steering motor provides a clear physical theoretical foundation for the estimation process, fully utilizing the known electrical parameters and mathematical model information of the motor to ensure the physical rationality of the estimation results. Estimating the back EMF as an extended state is equivalent to transforming the internal electromagnetic quantities that cannot be directly measured during motor operation into observable state variables, overcoming the limitation of traditional methods that struggle to directly obtain the back EMF. The extended state design also allows the observer to incorporate the effects of parameter changes, unmodeled dynamics, and external disturbances into the estimation scope, enhancing the estimation process's adaptability and robustness to changes in actual operating conditions. This online real-time estimation of the back EMF provides crucial input for subsequent rotor position and speed calculations, serving as a fundamental prerequisite for the implementation of the entire sensorless control method.

[0065] For example, the specific formula for the basic voltage balance model is as follows:

[0066]

[0067] in, For the stator resistance of the steering motor, For the stator inductance of the steering motor, , Let αβ be the actual current in the stationary coordinate system. , Let αβ be the actual voltage in the stationary coordinate system. , Let α be the actual back electromotive force in the αβ stationary coordinate system.

[0068] In one embodiment, the step of obtaining the estimated values ​​of the two stationary back electromotive forces in the αβ stationary coordinate system based on the three-phase current and three-phase voltage of the target steering motor using a pre-built back electromotive force extended state observer includes:

[0069] Clark transformations were performed on the three-phase current and three-phase voltage of the target steering motor to obtain the actual current signal and actual voltage signal in the αβ stationary coordinate system.

[0070] Based on the pre-constructed back EMF extended state observer, state estimation is performed using actual current and voltage signals to obtain the estimated current signal. Based on the deviation between the estimated current signal and the actual current signal, the estimated values ​​of the two stationary back EMFs in the αβ stationary coordinate system are calculated.

[0071] Clark transformation refers to the mathematical transformation that converts electrical quantities in a three-phase stationary coordinate system into electrical quantities in a two-phase stationary coordinate system. This transformation maps the three-phase physical quantities of the target steering motor onto the mutually perpendicular α and β axes. The actual current signal refers to the instantaneous two-phase current value in the αβ stationary coordinate system obtained after Clark transformation, and the actual voltage signal refers to the instantaneous two-phase voltage value in the αβ stationary coordinate system obtained after Clark transformation. The estimated current signal refers to the two-phase current value in the αβ stationary coordinate system reconstructed from the internal state of the back-EMF extended state observer after substituting the actual voltage signal as input for recursive calculation. The estimated stationary two-phase back-EMF values ​​refer to the α-axis and β-axis back-EMF components calculated in real time based on the deviation between the estimated current signal and the actual current signal, through the error correction mechanism within the back-EMF extended state observer.

[0072] In the specific implementation process, the instantaneous values ​​of the three-phase current of the target steering motor are first acquired through current sensors, and the instantaneous values ​​of the three-phase voltage are simultaneously obtained through voltage detection circuits. These analog signals are processed by conditioning circuits and then sent to the microcontroller's analog-to-digital converter to be converted into digital quantities. Inside the microcontroller, Clark transformation operations are performed on the three-phase current and three-phase voltage digital quantities respectively. That is, according to the constant amplitude or constant power transformation principle, the three-phase physical quantities are projected onto the αβ two-phase stationary coordinate system to obtain the actual current signal and the actual voltage signal. This transformation process eliminates the coupling relationship between the three-phase system, laying the foundation for subsequent state estimation based on the two-phase model. Subsequently, the actual voltage signal is sent as input to a pre-constructed back-EMF extended state observer. This observer contains a discretized state equation based on the steering motor voltage balance model. By performing recursive calculations in each control cycle, the internal state variables of the observer are updated in real time, and the estimated current signal is output. The estimated current signal represents the reconstruction result of the current value by the observer based on the motor model and the input voltage. The deviation between the estimated current signal and the actual current signal is calculated to obtain the current error value. According to the observer's design principles, this current error reflects the inconsistency between the model prediction and the actual measurement, one of the main reasons for which is the existence of back electromotive force (EMF). Therefore, this current error is corrected and fed back through the observer's built-in error gain matrix, continuously adjusting the observer's estimate of the back EMF, ultimately causing the observer to converge and output stable estimates of the two stationary back EMFs in the αβ stationary coordinate system.

[0073] In the actual operation of electric power steering systems, the motor often operates at low speeds and with frequent reversals. In such conditions, the current sampling signal is often mixed with inverter switching noise and electromagnetic interference. Converting the three-phase current to a two-phase current using Clark transform not only reduces the dimensionality and computational complexity of subsequent state estimation but also makes the noise distribution isotropic, facilitating unified processing by the observer. During the observer's recursive calculation, the deviation between the estimated current signal and the actual current signal is calculated in real time and used to correct the back EMF estimate, forming a closed-loop estimation structure. When the steering motor is stationary or at extremely low speeds, the back EMF signal is weak. At this time, the current deviation mainly originates from model errors and measurement noise. The observer can effectively suppress these interferences through its internal error gain matrix, preventing the back EMF estimate from diverging. As the motor starts to rotate or its speed changes, the back EMF gradually builds up, and the contribution of the back EMF to the current deviation increases. The observer can quickly respond to this change and update the back EMF estimate in a timely manner, ensuring it accurately reflects the motor's operating state. This bias-correction-based estimation method makes the entire back EMF acquisition process adaptive, enabling continuous and accurate estimation without relying on additional position sensors.

[0074] It should be noted that the Clark transform converts the three-phase electrical quantities into actual current and voltage signals in a two-phase stationary coordinate system, simplifying the mathematical model for subsequent state estimation, reducing computational complexity, and providing an efficient preprocessing procedure for real-time control. Based on the back-EMF extended state observer, state estimation is performed on the actual current and voltage signals to obtain the estimated current signal. The deviation from the measured current is then calculated, constructing a closed-loop feedback correction mechanism that makes the back-EMF estimation process convergent and adaptive. Using the deviation between the estimated and actual currents as the correction basis for back-EMF estimation effectively suppresses the influence of inaccurate model parameters and external disturbances on the estimation results, improving the tracking accuracy of the back-EMF estimate to the true value. The final stationary two-phase back-EMF estimate is obtained through continuous optimization under closed-loop correction, possessing high signal-to-noise ratio and stability, providing reliable basic data for the subsequent accurate calculation of rotor position and speed.

[0075] In one example, the three-phase current and three-phase voltage are subjected to Clark transformations respectively. The formulas for obtaining the actual current signal and actual voltage signal in the αβ stationary coordinate system are as follows:

[0076]

[0077]

[0078] in, , , This represents the actual three-phase current of the steering motor. , , This represents the actual three-phase voltage of the steering motor.

[0079] The specific formula for constructing the back-EMF extended state observer is as follows:

[0080]

[0081] In this formula, , This is the estimated current in the αβ stationary coordinate system of the previous period. , This is the estimated back electromotive force in the αβ stationary coordinate system of the previous period. The gain parameter for the back EMF extended state observer. , This is the derivative of the estimated current in the current αβ stationary coordinate system.

[0082] Calculate the back electromotive force of the motor using the estimated current and the actual current signal:

[0083]

[0084] In this formula, , This is the estimated current in the current αβ stationary coordinate system. , Let be the derivative of the estimated back electromotive force in the current αβ stationary coordinate system. This represents the gain parameter of the back-EMF extended state observer.

[0085] In one embodiment, the step of rotating two stationary back EMF estimates to obtain two rotated back EMF estimates and constructing a half-sine integral position error signal includes:

[0086] By performing a Park transformation on the estimated values ​​of the two opposing electromotive forces at rest, we obtain the estimated values ​​of the two opposing electromotive forces in the rotating dq coordinate system.

[0087] Calculate the estimated value of the back electromotive force amplitude based on the estimated values ​​of the two rotating back electromotive forces;

[0088] A half-sine integral type position error signal is constructed based on the specific axis components in the back EMF amplitude estimate and the back EMF estimates of the two rotating sides.

[0089] In this embodiment, the Park transformation refers to the mathematical transformation that maps electrical quantities in the αβ stationary coordinate system to the dq coordinate system, which rotates synchronously with the rotor. This transformation relies on real-time rotor position estimates and projects physical quantities from the stationary coordinate system to the rotating coordinate system by rotating the coordinate axes. The estimated back EMF values ​​for the two rotating axes refer to the two back EMF components along the d-axis and q-axis directions obtained in the dq rotating coordinate system after the Park transformation. The estimated back EMF amplitude refers to the magnitude of the composite back EMF vector calculated based on the estimated back EMF values ​​for the two rotating axes; it represents the overall strength of the back EMF signal. The specific axis component refers to the component in the estimated back EMF values ​​for the two rotating axes that is directly related to the rotor position error, typically the q-axis back EMF component. The half-sine integral type position error signal refers to the signal quantity characterizing the rotor position estimation deviation, constructed by normalizing the specific axis component based on the estimated back EMF amplitude and integrating the normalized signal within the integration interval corresponding to the half-wave width of the motor's electrical cycle.

[0090] In the specific implementation process, the rotor position estimate of the target steering motor is first obtained from the previous control cycle. This value can be the position result obtained from the previous calculation or the initial position set during system initialization. Using this rotor position estimate, a Park transformation matrix (i.e., a rotating coordinate transformation matrix) is constructed. The estimated values ​​of the two back electromotive forces (EMFs) in the stationary coordinate system αβ obtained in the previous stage, êα and êβ, are used as input to perform the Park transformation operation, resulting in estimated values ​​of the two rotating back EMFs in the rotating coordinate system dq. The essence of this transformation process is to map the back EMF signal, which contains sinusoidal fluctuation information in the stationary coordinate system, to a coordinate system that rotates synchronously with the rotor magnetic poles, giving the transformed direct-axis and quadrature-axis components clear physical meanings. Subsequently, a vector synthesis operation is performed on the estimated values ​​of the two rotating back EMFs, and the sum of their squares is calculated and then the square root is taken to obtain the estimated back EMF amplitude. This amplitude represents the overall intensity of the back EMF signal at the current moment and is approximately proportional to the motor speed. Next, the q-axis component is extracted from the back EMF estimates of the two rotating rotors. The q-axis component is then normalized using the back EMF amplitude estimate to obtain the normalized error index. The time length corresponding to the motor's electrical cycle is calculated based on the currently estimated rotor electrical angular velocity, and half of this time length is used as the width of the integration window. Within this integration window, the normalized error index is continuously accumulated or numerically integrated. When the integration window slides to the current moment, the result of the integration operation is the half-sine integral type position error signal. This construction method utilizes the back EMF amplitude to normalize the error signal, ensuring that the constructed error signal maintains relatively consistent gain characteristics as the motor speed changes.

[0091] In real-world operating scenarios of electric power steering systems, such as when a vehicle is turning in place or parking at low speed, the motor speed is low and fluctuates frequently. Converting the stationary back EMF to a rotating back EMF using the Park transform allows the error information related to rotor position to be concentrated in the q-axis component, achieving information decoupling and extraction. Since the back EMF amplitude is small and fluctuates with speed, directly using the q-axis component as the error signal would introduce speed-related gain changes, which is detrimental to the stable control of the subsequent speed loop. Therefore, normalizing the q-axis component using the estimated back EMF amplitude effectively eliminates the influence of speed changes on the error signal amplitude, ensuring that the constructed error signal reflects only the position deviation information and is decoupled from the speed. Furthermore, processing using a half-sine integral method, equivalent to averaging over half a wave of an electrical cycle, effectively filters out instantaneous disturbances caused by current sampling noise and inverter nonlinearity, resulting in a smoother output error signal that accurately reflects the average trend of the position deviation. This design, where the integration window width adaptively adjusts with the electrical cycle, ensures the consistency of the integration effect as the speed changes.

[0092] It should be noted that by converting the stationary back EMF to a rotating back EMF through the Park transform, the rotor position error information is decoupled and extracted, providing a clear physical basis for subsequent error quantization. Calculating the back EMF amplitude estimate based on the rotating back EMF and using this amplitude to normalize specific shaft components effectively eliminates the influence of speed variations on the error signal gain, decoupling the constructed position error signal from the motor operating speed and enhancing the universality and stability of the error signal. Using a half-sine integral to process the normalized signal, performing integral averaging within half the width of one electrical cycle, significantly suppresses the disturbances caused by instantaneous noise and back EMF harmonics on the error signal, improving the smoothness and signal-to-noise ratio of the error signal. This design, where the integral window adaptively adjusts with the electrical cycle, ensures that the error signal construction method maintains consistent filtering effects at different speeds. The final position error signal accurately reflects the average trend of the rotor position estimation deviation, providing a high-quality and low-fluctuation input signal for the subsequent fractional-order integral speed loop.

[0093] In one example, the specific formula for performing the Park transformation on two stationary opposite electromotive forces is as follows:

[0094]

[0095] in, This is the feedback estimate of the steering motor position from the previous cycle. , This is the estimated back electromotive force in the dq rotating coordinate system.

[0096] The specific formula for constructing a half-sine integral position error signal is as follows:

[0097]

[0098]

[0099] in, This is an estimate of the amplitude of the two opposing electromotive forces during rotation. It is a half-sine integral type position error signal.

[0100] In one embodiment, the formula for the target velocity estimate is:

[0101]

[0102] in, This represents the estimated target speed. This represents the proportional parameter of the fractional integral velocity loop. The integral parameters represent the fractional integral velocity loop. Denotes the Laplace field operator, This represents the fractional integral in the corresponding time domain. This indicates the adjustable fractional order. This indicates the position error signal.

[0103] This formula provides the mathematical expression for a fractional-order integral velocity loop, where the target velocity estimate is composed of the results of two parallel calculations. The first term is a proportional term, obtained by multiplying the proportional parameter by the position error signal, used for instantaneous amplification and response of the error signal. The second term is a fractional-order integral term, obtained by performing a fractional-order integration operation on the integration parameter and the position error signal, used for weighted historical accumulation processing of the error signal. The sum of these two terms serves as the final target velocity estimate output by the velocity loop. The fractional order in the formula is an adjustable parameter that determines how the fractional-order integration operation weights historical error information; different fractional orders correspond to different integral characteristics and frequency response characteristics.

[0104] By combining proportional and fractional integral terms in parallel, the velocity estimate can respond instantly to current errors while also eliminating steady-state deviations through the integral term, thus balancing dynamic response speed and steady-state accuracy. Introducing a fractional integral operator instead of the traditional integer integral allows the integral term to handle historical error information with varying weights, rather than simply adding them together. This enables flexible adjustment of the integral characteristics. When the fractional order is reduced, the weight of recent errors increases while the weight of long-term errors decreases, making the velocity loop more sensitive to changes in input errors and accelerating the response of the target velocity estimate during dynamic processes, allowing it to quickly follow changes in the actual rotational speed. When the fractional order is increased, the weighting range of historical errors by the integral term is wider, enhancing the smoothing effect and more effectively suppressing residual disturbances and noise in the input error signal. This results in a more stable target velocity estimate in steady state and reduced estimation drift. This ability to alter the speed loop response characteristics through a single parameter adjustment allows the same speed loop structure to adapt to the performance requirements of the steering motor under different operating conditions, avoiding the inherent problem of traditional fixed-structure integrators struggling to balance dynamic response speed and steady-state disturbance rejection performance. The resulting target speed estimate combines rapid following characteristics with steady-state smoothness, providing more accurate and reliable speed feedback information for subsequent rotor position estimation and closed-loop control.

[0105] In one embodiment, the fractional integral is defined using the Caputo operator, which is obtained by performing a convolution integral on the position error signal at historical time points with a power-law weighted function.

[0106] In this embodiment, the Caputo definition is a commonly used definition in fractional calculus theory. Mathematically, fractional integration under this definition represents a convolution integral of the input signal, with the kernel function being a weighting function of power-law form. Specifically, applying the Caputo definition to the position error signal is equivalent to performing a weighted summation of the position error signals from all historical moments at the current time according to a power-law function of time distance. The weights assigned to historical errors closer to the current time are power-law related to the weights assigned to historical errors further away from the current time. This weighting method is determined by the specific value of the fractional order.

[0107] Processing the position error signal using fractional integration under the Caputo definition to obtain the target speed estimate can produce a series of beneficial effects. Through the convolution integral form of the power-law weight function, the influence of historical error information on the current integral result is no longer of equal weight, but decays or enhances according to the power-law rule with time distance. This weighting method is more in line with the gradual change rule of the influence of historical states on the current state in actual physical systems. Since the power exponent in the power-law weight function is determined by the adjustable fractional order, by adjusting the value of the fractional order, the weight distribution characteristics of historical error information can be changed, realizing flexible control of the memory characteristics of the integrator. When the fractional order is adjusted downwards, the decay rate of the power-law weight function accelerates, and the integrator focuses more on recent error information and ignores long-term errors, enabling the speed estimate to quickly respond to transient changes in the input error. When the fractional order is adjusted upwards, the decay rate of the power-law weight function slows down, and the integrator still retains a certain weight for long-term error information, making the speed estimate smoother and having a stronger ability to suppress disturbances. This implementation method of fractional integration based on the Caputo definition provides a mathematically rigorous and physically meaningful processing means for the speed estimator, enabling the fractional integration type speed loop to achieve a continuous transition from first-order integration to near-integer-order integration through single-parameter adjustment, providing an accurate mathematical tool to adapt to the performance requirements of the steering motor under different working conditions. The finally obtained target speed estimate can quickly follow the change of the true speed during the dynamic process and effectively suppress the estimation drift caused by disturbances during the steady-state process, improving the estimation accuracy and operation stability of the entire sensorless control system under complex working conditions.

[0108] It can be understood that the definition of the α-order Caputo fractional derivative corresponding to the Caputo fractional integration is as follows:

[0109]

[0110] where n is a positive integer satisfying n - 1 < α < n, is the Gamma function, t is the current time, and τ is the integration variable; isthe target position estimate of the steering motor. The Gamma function, its standard definition is: for a complex number z satisfying Re(z) > 0, the Gamma function Γ(z) is a generalized integral in the form of Euler's second kind of integral, and the expression is:

[0111]

[0112] In fractional integration operations, the Gamma function, as the weight factor of the integral term, is used to determine the "smoothing weight" of fractional operations, ensuring the continuity and numerical rationality of fractional calculus operations. By adjusting the weights of the integral terms, a flexible weighting of the fractional derivative on "historical information" can be achieved.

[0113] In one embodiment, the formula for the target location estimate is:

[0114]

[0115] in, This represents the estimated location of the target. This represents the integer integral in the corresponding time domain.

[0116] This formula describes the transformation from a target velocity estimate to a target position estimate, where the target velocity estimate is used as an input quantity and then integrated by integer order to obtain the target position estimate. This mathematical expression reflects the fundamental physical relationship between rotational speed and angle; that is, the angle is the cumulative result of rotational speed over time. In a discrete control system, this is represented by multiplying the instantaneous velocity by the time step and then summing them up within each control cycle. The integer-order integral operator in the formula represents a mathematical processing method for continuously accumulating or summing the input signal, and is a standard mathematical tool for realizing the conversion from velocity information to position information.

[0117] A deterministic mathematical relationship between rotational speed and position is established through integer-order integration, giving the conversion process from speed information to position information a clear physical meaning and mathematical foundation, thus ensuring the physical rationality of the position estimation result. Integer-order integration inherently possesses smoothing and filtering characteristics, further suppressing residual high-frequency components in the input target speed estimate, making the output target position estimate smoother and more continuous, avoiding abnormal jumps in the position estimate caused by minor fluctuations in the speed estimate. Since the target speed estimate has already undergone fractional-order integral speed loop processing in the preceding steps, possessing good dynamic tracking performance and steady-state smoothness, the target position estimate obtained by integer-order integration inherits these advantages accordingly. It can quickly follow rotor position changes during dynamic processes and maintain the stability of the position estimate during steady-state processes, avoiding position drift. Integer-order integration has a simple structure, low computational load, and is easily discretized in microcontrollers, meeting the real-time calculation requirements under high control cycle frequencies without imposing an excessive computational burden on the control system. The final target position estimate provides accurate rotor position feedback for the vector control algorithm of the steering motor, which is a key prerequisite for the sensorless control system to achieve precise torque and angle control.

[0118] The fractional integral type steering motor position and speed estimation device provided in the embodiments of this application will be described below. The fractional integral type steering motor position and speed estimation device described below can be referred to in correspondence with the fractional integral type steering motor position and speed estimation method described above. Figure 2 As shown, this application provides a fractional integral type steering motor position and speed estimation device, the device comprising:

[0119] The stationary two-phase back electromotive force estimation module 201 is used to obtain the stationary two-phase back electromotive force estimation values ​​in the αβ stationary coordinate system based on the three-phase current and three-phase voltage of the target steering motor and using a pre-built back electromotive force extended state observer.

[0120] The position error signal construction module 202 is used to rotate the two stationary opposite electromotive force estimates to obtain the two rotated opposite electromotive force estimates, and construct a half-sine integral type position error signal.

[0121] The target velocity estimation module 203 is used to input the position error signal into a fractional-order integral velocity loop with an adjustable fractional order to obtain the target velocity estimate.

[0122] The target position estimation module 204 is used to perform integer-order integration on the target velocity estimation to obtain the target position estimation.

[0123] In one embodiment, the back EMF extended state observer is built based on the voltage balance model of the steering motor and is used to estimate the back EMF as an extended state in real time.

[0124] In one embodiment, the module 201 for determining the estimated values ​​of the two static back electromotive forces includes:

[0125] The actual electrical signal determination unit is used to perform Clark transformation on the three-phase current and three-phase voltage of the target steering motor to obtain the actual current signal and actual voltage signal in the αβ stationary coordinate system.

[0126] The unit for determining the estimated values ​​of the two stationary back electromotive forces is used to perform state estimation operations based on the pre-constructed back electromotive force extended state observer, using actual current signals and actual voltage signals to obtain estimated current signals. Based on the deviation between the estimated current signals and the actual current signals, the estimated values ​​of the two stationary back electromotive forces in the αβ stationary coordinate system are calculated.

[0127] In one embodiment, the position error signal construction module 202 includes:

[0128] The rotating two-opposite-electromotive force estimation unit is used to perform Park transformation on the stationary two-opposite-electromotive force estimation to obtain the rotating two-opposite-electromotive force estimation in the dq rotating coordinate system.

[0129] The back electromotive force amplitude estimation unit is used to calculate the back electromotive force amplitude estimate based on the back electromotive force estimates of the two rotating pairs;

[0130] The position error signal construction unit is used to construct a half-sine integral type position error signal based on a specific axis component in the back EMF amplitude estimate and the back EMF estimates of the two rotating sides.

[0131] In one embodiment, the formula for the target velocity estimate is:

[0132]

[0133] in, This represents the estimated target speed. This represents the proportional parameter of the fractional integral velocity loop. The integral parameters represent the fractional integral velocity loop. Denotes the Laplace field operator, This represents the fractional integral in the corresponding time domain. This indicates the adjustable fractional order. This indicates the position error signal.

[0134] In one embodiment, the fractional integral is defined using the Caputo operator, which is obtained by performing a convolution integral on the position error signal at historical time points with a power-law weighted function.

[0135] In one embodiment, the formula for the target location estimate is:

[0136]

[0137] in, This represents the estimated location of the target. This represents the integer integral in the corresponding time domain.

[0138] In one embodiment, this application also provides a storage medium storing computer-readable instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of the fractional integral type steering motor position and speed estimation method as described in any of the above embodiments.

[0139] In one embodiment, this application also provides a computer device storing computer-readable instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of the fractional integral type steering motor position and speed estimation method as described in any of the above embodiments.

[0140] Indicatively, such as Figure 3 As shown, Figure 3 This is a schematic diagram of the internal structure of a computer device 300 provided in an embodiment of this application. The computer device 300 can be provided as a server. (Refer to...) Figure 3 The computer device 300 includes a processing component 302, which further includes one or more processors, and memory resources represented by memory 301 for storing instructions, such as application programs, that can be executed by the processing component 302. The application programs stored in memory 301 may include one or more modules, each corresponding to a set of instructions. Furthermore, the processing component 302 is configured to execute instructions to perform the fractional integral type steering motor position and speed estimation method of any of the above embodiments.

[0141] The computer device 300 may also include a power supply component 303 configured to perform power management of the computer device 300, a wired or wireless network interface 304 configured to connect the computer device 300 to a network, and an input / output (I / O) interface 305. The computer device 300 may operate on an operating system stored in memory 301, such as Windows Server™, Mac OS X™, Unix™, Linux™, Free BSD™, or similar.

[0142] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0143] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element. In this document, "a," "an," "the," "the," and "its" may also include plural forms unless the context clearly indicates otherwise. "Multiple" refers to at least two, such as 2, 3, 5, or 8, etc. "And / or" includes any and all combinations of the related listed items.

[0144] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The various embodiments can be combined as needed, and the same or similar parts can be referred to each other.

[0145] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A fractional integral method for estimating the position and speed of a steering motor, characterized in that, The method includes: Based on the three-phase current and three-phase voltage of the target steering motor, the estimated values ​​of the two stationary back electromotive forces in the αβ stationary coordinate system are obtained by using a pre-constructed back electromotive force extended state observer. The two static opposite electromotive force estimates are rotated to obtain two rotated opposite electromotive force estimates, and a half-sine integral type position error signal is constructed. The position error signal is input into a fractional-order integral velocity loop with an adjustable fractional order to obtain the target velocity estimate. The target velocity estimate is obtained by performing an integer-order integral operation on the target velocity estimate.

2. The fractional integral type method for estimating the position and speed of a steering motor according to claim 1, characterized in that, The back EMF extended state observer is built based on the voltage balance model of the steering motor and is used to estimate the back EMF as an extended state in real time.

3. The fractional integral type method for estimating the position and speed of a steering motor according to claim 1, characterized in that, The step of obtaining the estimated values ​​of the two stationary back electromotive forces in the αβ stationary coordinate system based on the three-phase current and three-phase voltage of the target steering motor using a pre-constructed back electromotive force extended state observer includes: Clark transformations are performed on the three-phase current and three-phase voltage of the target steering motor to obtain the actual current signal and actual voltage signal in the αβ stationary coordinate system. Based on the pre-constructed back EMF extended state observer, state estimation is performed using the actual current signal and the actual voltage signal to obtain the estimated current signal. Based on the deviation between the estimated current signal and the actual current signal, the estimated values ​​of the two stationary back EMFs in the αβ stationary coordinate system are calculated.

4. The fractional integral type method for estimating the position and speed of a steering motor according to claim 1, characterized in that, The step of rotating the two stationary opposite electromotive force estimates to obtain two rotated opposite electromotive force estimates and constructing a half-sine integral type position error signal includes: The estimated values ​​of the two opposing electromotive forces at rest are subjected to Park transformation to obtain the estimated values ​​of the two opposing electromotive forces in the rotating dq coordinate system. Based on the estimated values ​​of the two opposing electromotive forces of rotation, calculate the estimated value of the back electromotive force amplitude; Based on the specific axis components in the estimated back EMF amplitude and the estimated back EMF of the two rotating sides, a half-sine integral type position error signal is constructed.

5. The fractional integral type method for estimating the position and speed of a steering motor according to claim 1, characterized in that, The formula for the target velocity estimate is: in, This represents the estimated target velocity. This represents the proportional parameter of the fractional integral velocity loop. This represents the integral parameter of the fractional-order integral velocity loop. Denotes the Laplace field operator, This represents the fractional integral in the corresponding time domain. This indicates the adjustable fractional order. This represents the position error signal.

6. The fractional integral type method for estimating the position and speed of a steering motor according to claim 5, characterized in that, The fractional integral is defined using the Caputo operator, which is a convolution integral of the position error signal at historical time points with a power-law weighted function.

7. The fractional integral type method for estimating the position and speed of a steering motor according to claim 5, characterized in that, The formula for the target location estimate is: in, This represents the estimated location of the target. This represents the integer integral in the corresponding time domain.

8. A fractional-order integral type steering motor position and speed estimation device, characterized in that, The device includes: The stationary two-phase back electromotive force estimation module is used to obtain the stationary two-phase back electromotive force estimates in the αβ stationary coordinate system based on the three-phase current and three-phase voltage of the target steering motor and using a pre-built back electromotive force extended state observer. The position error signal construction module is used to rotate the two stationary opposite electromotive force estimates to obtain two rotated opposite electromotive force estimates, and construct a half-sine integral type position error signal. The target velocity estimation module is used to input the position error signal into a fractional-order integral velocity loop with an adjustable fractional order to obtain the target velocity estimate. The target position estimation module is used to perform integer-order integration on the target velocity estimation to obtain the target position estimation.

9. A storage medium, characterized in that: The storage medium stores computer-readable instructions that, when executed by one or more processors, cause the one or more processors to perform the steps of the fractional integral type steering motor position and speed estimation method as described in any one of claims 1 to 7.

10. A computer device, characterized in that, include: One or more processors, and memory; The memory stores computer-readable instructions that, when executed by the one or more processors, perform the steps of the fractional integral type steering motor position and speed estimation method as described in any one of claims 1 to 7.

Images