Single current sensor based speed re-throw method and system based on line voltage vector injection

Abstract

The application discloses a single current sensor speed re-throwing method and system based on line voltage vector injection, which is based on the defined line voltage vector injection mode, periodically turns on two-phase bridge arms of an inverter, collects the current during the motor inertial operation by using a single current sensor, extracts the line back electromotive force, thereby realizes fast and accurate motor rotor position and speed information acquisition, and completes the seamless smooth cut-in from the motor inertial operation to the non-inductive control. The application is suitable for the motor driving control system of a three-phase two-level inverter, and only needs to inject at least three line voltage vectors to realize the rotor position and speed information acquisition during the motor inertial operation. Meanwhile, the application does not require the inertia size of the motor, is suitable for common single current sensor driving control systems, and can realize the speed re-throwing in a wide speed range of the motor.

Description

Technical Field

[0001] This invention belongs to the field of permanent magnet synchronous motor drive control, and particularly relates to a single current sensor belt speed re-switching method and system based on line voltage vector injection. Background Technology

[0002] In sensorless applications, permanent magnet synchronous motor drive systems may require speed-on reconfiguration in special load scenarios such as fans and pumps, as well as during short-term power outages. During these situations, the motor is in an inertial running state; abruptly restarting it can lead to a failure to match rotor position and speed information, potentially causing restart failure or even a large current surge that could result in inverter failure. To address these issues, academia and industry have conducted research on speed-on reconfiguration. Existing solutions based on multi-current sensors include terminal voltage measurement, zero-current vector closed-loop, zero-voltage vector pulse, virtual resistance, and high-frequency injection methods. Among the various methods, the terminal voltage measurement method is relatively simple. It calculates the back electromotive force (EMF) of the motor during inertial operation by sampling the three-phase terminal voltages, thereby extracting the rotor position and speed information. However, it requires additional voltage sampling of the terminal voltages. The zero-current vector closed-loop method sets the current loop setpoint to zero, thereby matching the inverter's output voltage with the back EMF, and then observes the inverter voltage setpoint to achieve rotor position and speed observation. However, this method may trigger overcurrent during the adjustment process. The zero-voltage vector pulse method applies a zero vector to the inverter, turning on all upper or lower bridge arm switches to induce a momentary short circuit in the motor, thus utilizing... The short-circuit current is excited by back electromotive force, and the rotor position and speed are extracted by sampling the magnitude of the short-circuit current. In addition, this method can be further classified according to the number of zero-voltage vector pulses applied. The virtual resistance method introduces a virtual resistor in the loop and adjusts the size of the virtual resistor to allow a small stator current to flow, and then estimates the rotor position and speed of the motor based on the current. However, the observation accuracy of this method is affected by the initial speed. The high-frequency injection method injects high-frequency components in different coordinate systems and extracts the high-frequency response signal in the motor to obtain rotor position information. However, this method requires magnetic pole identification.

[0003] The aforementioned solution cannot be applied to belt-speed re-start applications of permanent magnet synchronous motors based on a single current sensor because its current detection method differs, making it impossible to simultaneously detect three-phase current information and thus estimate the current rotor position and speed. Currently, due to its advantages of low cost, small size, and high system reliability, belt-speed re-start solutions based on a single current sensor are widely used. Therefore, the design of a belt-speed re-start solution based on a single current sensor has significant practical implications. Summary of the Invention

[0004] To address the technical problems mentioned above, this invention discloses a single current sensor belt-speed re-switching method and system based on line voltage vector injection.

[0005] This invention is implemented according to the following technical solution:

[0006] In a first aspect, the present invention provides a single-current sensor-based speed-reconnection method based on line voltage vector injection, applicable to a permanent magnet synchronous motor system driven by a three-phase two-level inverter, the method comprising: Define the switching state of the inverter, allowing only two phase arms to perform complementary switching actions, forcing all upper and lower transistors of the remaining phase arm to turn off, and generating a set of line voltage vectors including a single-phase floating phase. Based on the installation topology of the current sensor, vectors that can form an effective sampling path are selected from the line voltage vectors and used as the injection vector group; Within one injection cycle, vectors of the vector group are injected into the inverter sequentially, and before switching to the next vector, the motor winding current is allowed to decay to zero to ensure that the initial current value when each vector is applied is zero. Collect current data during the action of each injected line voltage vector, and combine the symmetry of motor winding parameters to establish an equivalent circuit equation that ignores stator resistance voltage drop; By performing pairwise operations on the equivalent circuit equations corresponding to different injected line voltage vectors and eliminating the bus voltage term, a linear relationship between the line back electromotive force and the rate of change of current is obtained. The re-start factor is defined as the difference in the current slope in the linear relationship between the line back electromotive force and the rate of change of current; the re-start factor corresponds to the line back electromotive force and includes the speed and rotor position information required for belt speed re-start. After normalizing the re-entry factor estimation results, the results are fed into the quadrature phase-locked loop. The output of the phase-locked loop is compensated based on the phase offset between the line back electromotive force and the rotor flux linkage to obtain the rotor position and speed information of the motor during inertial operation. The rotor position and speed information are transmitted to the sensorless control observer, the inverter is restarted, and the motor is seamlessly switched from inertial operation to sensorless control.

[0007] In one embodiment, the line voltage vector of the phase floating includes 12 line voltage vectors, specifically: (00X), (01X), (10X), (11X), (0X0), (0X1), (1X0), (1X1), (X00), (X01), (X10), (X11), where "X" indicates that the upper and lower bridge arms of the phase are completely turned off.

[0008] In one embodiment, when the current sensor is a single-branch current sampling topology, the injection vector group includes four vectors (X10), (X01), (01X), and (0X1), which are defined as LV1(X10), LV2(X01), LV3(01X), and LV4(0X1), respectively.

[0009] In one embodiment, the reinvestment factor is a linearly symmetric three-phase reinvestment factor. , , The specific expression is: , In the formula, The electric angular velocity of the motor. It is a permanent magnet flux linkage. For rotor electrical angle, For phase inductance, For mutual induction, , , , These are the sampled values ​​of the motor phase current under the action of the corresponding line voltage vector; The three-phase re-entry factor is first subjected to Clark transformation and then fed into a quadrature phase-locked loop. The feedback angle inside the phase-locked loop needs to be superimposed with a phase compensation amount of π / 6.

[0010] In one implementation, the PI parameters of the quadrature phase-locked loop are set according to the optimal damping ratio of the second-order system, satisfying: , in, The proportional gain of the quadrature phase-locked loop. The integral gain of the quadrature phase-locked loop (PLL) is... It is the undamped natural frequency.

[0011] In one implementation, by optimizing the injection line voltage vector combination, the injection vector group is reduced to three vectors: LV1 (X10), LV2 (X01), and LV3 (01X). The reinvestment factor is orthogonally symmetrical. , The specific expression is: , In the formula, The electric angular velocity of the motor. It is a permanent magnet flux linkage. For rotor electrical angle, For phase inductance, For mutual induction, , , These are the sampled values ​​of the motor phase current under the action of the injected vector group; The re-projection factor does not require Clark transformation and is directly fed into the quadrature phase-locked loop. The feedback angle inside the phase-locked loop needs to be superimposed with a phase compensation amount of π / 3.

[0012] In one embodiment, when the current sensor is a DC bus current sampling topology, the injection vector group includes six vectors (01X), (10X), (0X1), (1X0), (X01), and (X10), which are defined as LV1(01X), LV2(10X), LV3(0X1), LV4(1X0), LV5(X01), and LV6(X10), respectively.

[0013] In one implementation, by optimizing the injection vector combination, the injection vector group reduces the three vectors LV2 (10X), LV5 (X01), and LV6 (X10); The reinvestment factor is orthogonally symmetrical. , The specific expression is: , In the formula, The electric angular velocity of the motor. It is a permanent magnet flux linkage. For rotor electrical angle, For phase inductance, For mutual induction, , , These are the sampled values ​​of the motor phase current under the action of the injected vector group; The re-projection factor does not require Clark transformation and is directly fed into the quadrature phase-locked loop. The feedback angle inside the phase-locked loop needs to be superimposed with a phase compensation amount of π / 3.

[0014] In one implementation, the current sampling time during each line voltage vector action is set at the end of the vector injection cycle, with a preset sampling hold time reserved.

[0015] This invention also provides a single-current sensor-based speed-reset system based on line voltage vector injection, suitable for a permanent magnet synchronous motor system driven by a three-phase two-level inverter. The system includes: Vector generation module: used to define the switching state of the inverter, only allowing two phase arms to perform complementary switching actions, forcing all upper and lower transistors of the remaining phase arm to be turned off, and generating a line voltage vector including a single-phase floating phase. Vector filtering module: used to filter out vectors that can form an effective sampling path from the line voltage vectors based on the installation position of the single current sensor, and construct an injection vector group; Injection control module: used to sequentially inject the vectors in the injection vector group into the inverter within one injection cycle, and wait for the motor winding current to decay to zero before switching to the next vector, to ensure that the initial current value when each vector is applied is zero; Equation Establishment Module: Used to collect current data during the action of each injected line voltage vector, and combine the symmetry of motor winding parameters to establish an equivalent circuit equation that ignores the stator resistance voltage drop; Differential operation module: used to perform differential operation on the equivalent circuit equations corresponding to different injected line voltage vectors, eliminate the bus voltage term, and obtain the observation expression in which the line back electromotive force is linearly related to the rate of change of current; Re-start factor calculation module: used to define the re-start factor as the difference in current slope in the observation expression. The re-start factor is in phase with the line back electromotive force and includes the speed and rotor position information required for belt speed re-start. Phase-locked loop (PLL) calculation module: This module normalizes the re-entry factor and feeds it into the quadrature PLL. It compensates for the PLL output based on the inherent phase offset between the line back electromotive force and the rotor flux linkage, and calculates the rotor position and speed information of the motor during inertial operation. Observer initialization module: used to assign the rotor position and speed information to the initial state of the sensorless control observer, restart the inverter, and complete the seamless transition of the motor from inertial operation to sensorless control.

[0016] Beneficial effects of this invention:

[0017] 1. This invention differs from traditional voltage vectors. The line voltage vector acts on the switching state of two phases, and the current generated by the line back electromotive force can effectively extract the line back electromotive force information.

[0018] 2. This invention equivalences and simplifies the circuits of the inverter and permanent magnet synchronous motor under line voltage vector action, and can obtain the equivalent circuit equations, which facilitates the analysis of circuit and motor characteristics.

[0019] 3. This invention, by performing pairwise operations on the simplified equivalent circuit equations during the defined line voltage vector action period, can obtain a result excluding the bus voltage V. dc The system of equations for the term facilitates the extraction of back electromotive force information.

[0020] 4. This invention proposes a method for defining a re-spinning factor. By defining the difference in current slopes that include information on the symmetry of the line back electromotive force as the re-spinning factor and sending it into a phase-locked loop for processing, the rotor position and speed information of the motor during inertial operation can be obtained.

[0021] 5. This invention is applicable not only to single-branch current sampling topologies but also to other common single-current sensor sampling topologies, such as DC bus current sampling topologies. Furthermore, this invention provides a design process for a line voltage vector injection method with speed-based re-switching for DC bus current sampling topologies. Attached Figure Description

[0022] The accompanying drawings, as part of this invention, are provided to further illustrate the invention. The illustrative embodiments and descriptions of the invention are used to explain the invention, but do not constitute an undue limitation thereof. Clearly, the drawings described below are merely some embodiments, and those skilled in the art can obtain other drawings based on these drawings without any creative effort.

[0023] Figure 1 This is a flowchart of a single current sensor speed-based re-switching method based on line voltage vector injection, provided in an embodiment of the present invention.

[0024] Figure 2 This is the current path provided in one embodiment of the present invention when the line voltage vector is applied, wherein... Figure 2 In diagram (a), the current path is shown when the line voltage vector (10X) is applied. Figure 2 In diagram (b), the current path is shown when the line voltage vector (1X0) is applied.

[0025] Figure 3 This is a current path provided by an embodiment of the present invention when a line voltage vector is applied; wherein Figure 3 In diagram (a), the current path is shown when the line voltage vector LV1(X10) is applied. Figure 3 In diagram (b), the current path is shown when the line voltage vector LV2(X01) is applied. Figure 3 In diagram (c), the current path is shown when the line voltage vector LV3(01X) is applied. Figure 3 In the middle (d), the current path is when the line voltage vector LV4 (0X1) is applied;

[0026] Figure 4 This is the equivalent circuit provided in one embodiment of the present invention when the line voltage vector LV3(01X) is in operation;

[0027] Figure 5 This is a phase-locked loop used for belt-speed re-switching provided in one embodiment of the present invention;

[0028] Figure 6 The vector injection sequence is provided in one embodiment of the proposed belt speed re-throw method of the present invention;

[0029] Figure 7 This is a current waveform during vector injection provided in one embodiment of the present invention, wherein... Figure 7 In diagram (a), the three-phase current waveforms are shown. Figure 7 (b) shows the waveform of the upper branch current;

[0030] Figure 8 The belt speed re-engagement process provided in one embodiment of the present invention is an initial speed of 1000 r / min under no-load conditions and a restart speed of 1000 r / min: wherein, Figure 8 In the middle (a), the actual rotational speed waveform is shown. Figure 8 (b) shows the three-phase current waveform. Figure 8 (c) shows the voltage waveform of line AB;

[0031] Figure 9 This is a normalized reprojection factor waveform in a stationary coordinate system provided in one embodiment of the present invention;

[0032] Figure 10 The rotor position estimation effect of the pulse injection process provided in one embodiment of the present invention;

[0033] Figure 11 The rotational speed estimation effect of the pulse injection process provided in one embodiment of the present invention;

[0034] Figure 12 This is a diagram illustrating the effect of the method proposed in an embodiment of the present invention under inertial operation at 300 rpm (20% of rated speed). Figure 12 (a) shows the rotor position estimation result. Figure 12 (b) shows the speed estimation result;

[0035] Figure 13 This is a diagram illustrating the effect of the method proposed in an embodiment of the present invention under inertial operation at 800 rpm (53% of rated speed). Figure 13 (a) shows the rotor position estimation result. Figure 13 (b) shows the speed estimation result;

[0036] Figure 14 This is a diagram illustrating the effect of the method proposed in an embodiment of the present invention under inertial operation at 1300 rpm (87% of rated speed). Figure 14 (a) shows the rotor position estimation result. Figure 14 (b) shows the speed estimation result;

[0037] Figure 15 This is a diagram illustrating the effect of belt-driven restart when the wind turbine is inertly running at 300 rpm under load, according to one embodiment of the present invention. Figure 15 In (a), the target speed is 500 rpm. Figure 15 (b) represents the target speed of 100 rpm. Figure 15 (c) shows the C-phase current waveform during pulse injection;

[0038] Figure 16 This is a diagram illustrating the effect of belt-driven re-energization under a wind turbine load of 800 rpm during inertial operation, as provided in one embodiment of the present invention: Figure 16 In (a), the target speed is 1000 rpm. Figure 16 (b) represents the target speed of 600 rpm;

[0039] Figure 17 This is a diagram illustrating the effect of belt-driven re-energization under a wind turbine load of 1300 rpm during inertial operation, as provided in one embodiment of the present invention. Figure 17 In (a), the target speed is 1500 rpm. Figure 17 In the middle (b), the target speed is 1100 rpm.

[0040] It should be noted that these accompanying drawings and textual descriptions are not intended to limit the scope of the invention in any way, but rather to illustrate the concept of the invention to those skilled in the art by referring to specific embodiments. Detailed Implementation

[0041] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments will be clearly and completely described below with reference to the accompanying drawings. The following embodiments are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0042] This invention provides a single-current sensor on-line speed re-switching method based on line voltage vector injection, such as... Figure 1 As shown, it includes the following steps:

[0043] S100: Defines the switching state of the inverter, allowing only two phase arms to perform complementary switching actions, forcing all upper and lower transistors of the remaining phase arm to turn off, and generating a set of line voltage vectors including a single-phase floating phase.

[0044] Furthermore, the single-phase floating line voltage vector includes 12 types of line voltage vectors, specifically: (00X), (01X), (10X), (11X), (0X0), (0X1), (1X0), (1X1), (X00), (X01), (X10), (X11), where “X” indicates that the upper and lower bridge arms of that phase are completely turned off.

[0045] For single-branch current sampling topologies, the injection of space voltage vectors acting on three-phase switching states cannot extract the back electromotive force. Therefore, this invention proposes a method for injecting line voltage vectors to extract line back electromotive force information, which can achieve fast and accurate acquisition of rotor position and speed information.

[0046] Specifically, unlike traditional voltage vectors, line voltage vectors act on the switching states of two phases, while the upper and lower arms of the other phase are completely off. For example, in the line voltage vector (10X), "1" represents the upper arm of the phase being on, "0" represents the lower arm of the phase being on, and "X" represents all arms of the phase being off. Thus, (10X) represents the upper arm of phase A being on, the lower arm of phase B being on, and all arms of phase C being off. Table 1 below summarizes 12 types of line voltage vectors.

[0047] Table 1

[0048]

[0049] S200: Based on the installation topology of the current sensor, select vectors from the line voltage vectors that can form an effective sampling path and use them as the injection vector group.

[0050] Due to the characteristics of a single-branch current sampling topology, current cannot be sampled during all periods of line voltage vector action. For example... Figure 2 As shown, during the period when the line voltage vectors (10X) and (1X0) are in effect, the current does not pass through the branch between the upper arms of the AB phases, so the effective current cannot be collected.

[0051] S300: Within one injection cycle, vectors of the vector group are injected into the inverter sequentially, and before switching to the next vector, the motor winding current is allowed to decay to zero to ensure that the initial current value when each vector is applied is zero.

[0052] S400: Collects current data during the action of each injected line voltage vector, and establishes an equivalent circuit equation that ignores the stator resistance voltage drop by combining the symmetry of the motor winding parameters.

[0053] Figure 3 This represents the current path when the line voltage vector is applied. Figure 3 In diagram (a), the current path is shown when the line voltage vector LV1(X10) is applied. Figure 3 In diagram (b), the current path is shown when the line voltage vector LV2(X01) is applied. Figure 3 In diagram (c), the current path is shown when the line voltage vector LV3(01X) is applied. Figure 3 In the middle (d), the current path is the current path when the line voltage vector LV4 (0X1) is applied.

[0054] Specifically, with Figure 3 Taking (c) as an example, the circuit during the operation of the line voltage vector LV3(01X) is simplified, and the following is obtained: Figure 3 The equivalent circuit shown is as follows, in which Bus voltage , This is the phase resistance of the motor. , For motor phase inductance, , This refers to the mutual inductance between the A and B phase windings of the motor. It is the line back electromotive force between the BA phases of the motor. During inertial operation, its phase changes with the position of the motor rotor, and its amplitude decreases as the motor speed decreases.

[0055] Based on its equivalent circuit, the equivalent circuit equation can be written as shown in the following equation.

[0056] (1)

[0057] In the formula, where Bus voltage , This is the phase current of the motor. , This is the phase resistance of the motor. , For motor phase inductance, , This refers to the mutual inductance between the A and B phase windings of the motor. This is the back electromotive force between phases BA of the motor.

[0058] Meanwhile, based on the characteristics of motor parameters and its equivalent circuit, the following simplified conditions can be derived.

[0059] (2)

[0060] In the formula, , This is the phase current of the motor. , This is the phase resistance of the motor. , For motor phase inductance, , This refers to the mutual inductance between the A and B phase windings of the motor. For phase inductance, For mutual induction.

[0061] Furthermore, since the motor resistance is small and the injected current is small, the voltage drop across the resistance can be ignored. Therefore, the equivalent circuit equation of formula (1) can be further simplified to the following formula, where Represents the injected line voltage vector The upper branch current was collected at that time.

[0062] (3)

[0063] When the current sensor is a single-branch current sampling topology, the injection vector group contains four vectors: (X10), (X01), (01X), and (0X1), which are defined as LV1(X10), LV2(X01), LV3(01X), and LV4(0X1), respectively.

[0064] The following table summarizes Figure 4 The equivalent circuit equations for each type of line voltage vector action.

[0065] Table 2

[0066]

[0067] S500: Perform pairwise operations on the equivalent circuit equations corresponding to different injected line voltage vectors, eliminate the bus voltage term, and obtain the linear relationship between the line back electromotive force and the rate of change of current.

[0068] By observing the equivalent circuit equations in Table 2, it can be seen that the bus voltage The extraction of back EMF information is affected by several factors. Meanwhile, the ideal motor line voltage formula is shown below, while the back EMF information in Table 2 requires further processing.

[0069] (4)

[0070] In the formula, , , This is the back electromotive force between two phases of the motor. The electric angular velocity of the motor. It is a permanent magnet flux linkage. The rotor electrical angle.

[0071] Therefore, according to the basic rules of vector operations, operations are performed on the equivalent circuit equations in the table. Subtracting (2) from (4) in the table and rearranging the equations, the formula can be eliminated. From this, we can derive the formula for the back electromotive force term of line AB in the following equation, and the other two equations are derived in a similar manner.

[0072] (5)

[0073] In the formula, , , This is the back electromotive force between two phases of the motor. For phase inductance, For mutual induction, , , , These are the sampled values ​​of the motor phase current under the action of the corresponding line voltage vector;

[0074] S600: The re-start factor is defined as the difference in the current slope in the linear relationship between the line back electromotive force and the rate of change of current; the re-start factor corresponds to the line back electromotive force and includes the speed and rotor position information required for belt speed re-start.

[0075] The back EMF term in equation (5) corresponds to the ideal back EMF term in the formula, which includes the speed and rotor position information required for belt-driven re-start. For ease of representation, the re-start factor (Rf) is defined as the difference in current slope in the above formula, and the linearly symmetrical three-phase re-start factor shown in the following formula can be obtained. , , :

[0076] (6)

[0077] In the formula, The electric angular velocity of the motor. It is a permanent magnet flux linkage. For rotor electrical angle, For phase inductance, For mutual induction, , , , These are the sampled values ​​of the motor phase current under the action of the corresponding line voltage vector.

[0078] The three-phase re-switching factor is first subjected to Clark transformation and then fed into the quadrature phase-locked loop. The feedback angle inside the phase-locked loop needs to be superimposed with a phase compensation amount of π / 6.

[0079] S700: After normalizing the re-entry factor estimation results, the S700 is fed into the quadrature phase-locked loop (PLL). The PLL output is compensated based on the phase offset between the line back electromotive force and the rotor flux linkage to obtain the rotor position and speed information of the motor during inertial operation.

[0080] The above re-entry factor estimation results are fed into a quadrature phase-locked loop (PLL). Since the estimated quantity is the line back electromotive force, it is necessary to slightly modify the traditional phase-locked loop that targets phasors. Its structure diagram is shown in Figure 5. The three-phase re-entry factor is then processed. , , Perform coordinate transformation to obtain , Then normalize to obtain , Finally, the estimated rotor position and speed information are obtained through phase-locked loop processing. Figure 5 In this case, δ is approximately equal to the rotor position estimation error. Note that the angle value of the internal feedback needs to be increased by π / 6.

[0081] Furthermore, since the traditional quadrature phase-locked loop is a second-order control system, the damping ratio can be set to 0.707, therefore the PI parameter can be set as follows:

[0082] (7)

[0083] in, The proportional gain of the quadrature phase-locked loop. The integral gain of the quadrature phase-locked loop (PLL) is... It is the undamped natural frequency.

[0084] S800: Transmits rotor position and speed information to the sensorless control observer, restarts the inverter, and completes the seamless transition of the motor from inertial operation to sensorless control.

[0085] In practice, the following methods are used: Figure 6 The sequence shown injects pulses into the inertial motor system, where Represents the injection cycle of speed-up reinvestment. This represents the actual injection time of the pulse in the injection cycle, with the remaining time waiting for the inductor current to decay to zero. The sampling time for the upper branch current is placed at the end of the actual pulse injection, with a certain sampling and holding time.

[0086] The three-phase current waveforms during the pulse injection process are as follows: Figure 7 As shown in (a), the two-phase currents are symmetrical, and the one-phase current is zero. The actual current waveforms differ due to factors such as the phase of the back electromotive force and the injection time. The upper branch current during the pulse injection process is as follows: Figure 7 As shown in (b), part of the waveform coincides with the phase current waveform.

[0087] Once the estimated rotational angle and speed have stabilized, preparations can be made for belt speed re-energization and inverter restart. For sensorless control, the observer needs initial value settings before belt speed re-energization to achieve seamless integration of speed and angle estimations. First, the estimated speed and angle during pulse injection are assigned to the observer's estimated speed and angle for speed-current loop calculations and coordinate transformations. In practical applications, there are often ramp speed change modules and low-pass filtering modules for the observer's speed; therefore, the current value of the ramp speed change and the filtering result from the previous cycle also need to be assigned before belt speed re-energization.

[0088] To further reduce the number of injection line voltage vectors in a full injection cycle, this invention optimizes the combination of injection line voltage vectors, reducing the number to three.

[0089] Furthermore, by injecting line voltage vectors LV1(X10), LV2(X01), and LV3(X01X), and performing pairwise operations on the simplified equivalent circuit equations during the action of the three line voltage vectors, a result excluding the bus voltage can be obtained. The following formula, which is convenient for extracting line back electromotive force information, is also applicable:

[0090] (8)

[0091] Based on the above formula, a new re-entry factor is defined to represent the difference in current slopes in the above formula, resulting in the orthogonally symmetrical re-entry factor shown in the following formula. , :

[0092] (9)

[0093] In the formula, The electric angular velocity of the motor. It is a permanent magnet flux linkage. For rotor electrical angle, For phase inductance, For mutual induction, , , These are the sampled values ​​of the motor phase current under the action of the injected vector group.

[0094] The newly defined re-entry factor estimation results first need to be normalized. Without Clark transformation, they can be fed into the quadrature phase-locked loop (PLL) for processing to obtain rotor position and speed information. Since the estimated quantity is the linear back electromotive force, the traditional phasor-based PLL needs to be slightly modified; that is, the internal feedback angle quantity needs to be increased by π / 3.

[0095] The design concept of this invention is not only applicable to single-branch current sampling topologies, but also to other common single-current sensor sampling topologies, such as DC bus current sampling topologies.

[0096] For the DC bus current sampling topology, based on its current sampling characteristics, there are also 12 identical line voltage vectors. However, not all line voltage vectors can collect effective current during their operation. When the current sensor is placed on the lower side of the DC bus, only six line voltage vectors (01X), (10X), (0X1), (1X0), (X01), and (X10) can collect effective current during their operation; the other six line voltage vectors cannot collect effective current information. Therefore, six new line voltage vectors, distinct from the single-branch current sampling topology, are defined as LV1(01X), LV2(10X), LV3(0X1), LV4(1X0), LV5(X01), and LV6(X10).

[0097] Following the same analytical approach, the injection line voltage vector combination was further optimized to achieve the goal of injecting only line voltage vectors LV2(10X), LV5(X01), and LV6(X10) within a full injection cycle. By performing pairwise operations on the simplified equivalent circuit equations during the action of the three line voltage vectors, it is possible to obtain the result excluding the bus voltage. The following formula, which is convenient for extracting line back electromotive force information, is also applicable:

[0098] (10)

[0099] Redefining the re-entry factor to represent the difference in current slopes in the above equation yields the orthogonally symmetrical re-entry factor shown in the following equation. , :

[0100] (11)

[0101] In the formula, The electric angular velocity of the motor. It is a permanent magnet flux linkage. For rotor electrical angle, For phase inductance, For mutual induction, , , These are the sampled values ​​of the motor phase current under the action of the injected vector group.

[0102] The newly defined re-entry factor estimation results first need to be normalized. Without Clark transformation, they can be fed into the quadrature phase-locked loop (PLL) for processing to obtain rotor position and speed information. Since the estimated quantity is a linear back electromotive force, the traditional PLL that targets phasors needs to be slightly modified; that is, the internal feedback angle quantity needs to be increased by π / 3.

[0103] Furthermore, the current sampling time during each line voltage vector action is set at the end of the vector injection cycle, and a sampling hold time of preset duration is reserved.

[0104] The following is an embodiment of a single current sensor belt-speed re-switching system based on line voltage vector injection according to the present invention, which can be used to execute an embodiment of a single current sensor belt-speed re-switching method based on line voltage vector injection according to the present invention. For details not disclosed in the embodiment of the single current sensor belt-speed re-switching system based on line voltage vector injection according to the present invention, please refer to the embodiment of the single current sensor belt-speed re-switching method based on line voltage vector injection according to the present invention.

[0105] In one embodiment, a single-current sensor-based speed-reconnection system based on line voltage vector injection is proposed, applicable to a permanent magnet synchronous motor system driven by a three-phase two-level inverter. The system includes:

[0106] Vector generation module: used to define the switching state of the inverter, only allowing two phase arms to perform complementary switching actions, forcing all upper and lower transistors of the remaining phase arm to be turned off, and generating a line voltage vector including a single-phase floating phase.

[0107] Vector filtering module: used to filter out vectors that can form an effective sampling path from the line voltage vectors based on the installation position of the single current sensor, and construct an injection vector group;

[0108] Injection control module: used to sequentially inject the vectors in the injection vector group into the inverter within one injection cycle, and wait for the motor winding current to decay to zero before switching to the next vector, to ensure that the initial current value when each vector is applied is zero;

[0109] Equation Establishment Module: Used to collect current data during the action of each injected line voltage vector, and combine the symmetry of motor winding parameters to establish an equivalent circuit equation that ignores the stator resistance voltage drop;

[0110] Differential operation module: used to perform differential operation on the equivalent circuit equations corresponding to different injected line voltage vectors, eliminate the bus voltage term, and obtain the observation expression in which the line back electromotive force is linearly related to the rate of change of current;

[0111] Re-start factor calculation module: used to define the re-start factor as the difference in current slope in the observed expression. The re-start factor is in phase with the line back electromotive force and includes the speed and rotor position information required for belt speed re-start.

[0112] Phase-locked loop (PLL) calculation module: This module normalizes the re-entry factor and feeds it into the quadrature PLL. It compensates for the PLL output based on the inherent phase offset between the line back electromotive force and the rotor flux linkage, and calculates the rotor position and speed information of the motor during inertial operation.

[0113] Observer initialization module: used to assign the rotor position and speed information to the initial state of the sensorless control observer, restart the inverter, and complete the seamless transition of the motor from inertial operation to sensorless control.

[0114] The functional modules in this embodiment of the invention can be integrated into one processing module, or each unit can exist as a separate physical entity, or two or more units can be integrated into one module. The integrated module can be implemented in hardware or as a software functional module.

[0115] For the description of the line voltage vector injection speed-reset method for other single current sensors, please refer to the description of the above embodiments, and it will not be repeated here.

[0116] To verify the effectiveness and reliability of the proposed single-current sensor speed-based re-switching method based on line voltage vector injection, a simulation was built in MATLAB / Simulink, simulating a wind turbine load with a bus voltage of 12V and a switching frequency of 20kHz.

[0117] Taking a single-branch current sampling topology and a four-wire voltage vector injection method as an example, to demonstrate the pulse injection process and the details of rotor position and speed estimation, in the simulation, the simulated motor enters inertial operation at an initial speed of 1000 rpm under no-load conditions at 0.1s, and the inverter is turned off. During this period of 0.1-0.12s, the motor is completely in an inertial operation state. At 0.12s, pulses are injected into the inertial motor system according to the proposed belt-speed re-start method, and the rotor position and speed are estimated simultaneously. Finally, at 0.2s, the estimated rotor position and speed information are used for belt-speed re-start, controlling the motor system to smoothly restart to 1000 rpm.

[0118] Figure 7 The actual rotational speed waveform, three-phase current waveform, and AB line voltage waveform are displayed throughout the entire process. Figure 7 The actual speed waveform shows that after entering inertial operation at 0.1s, the motor decreases at an extremely low speed due to the relatively ideal simulation system. However, after 0.12s, due to the torque effect of the injected pulse and the resulting motor current, the motor begins to decrease at a large, fluctuating slope. It restarts at a relatively accurate rotor position and speed after 0.2s, achieving a rapid and stable speed recovery. Figure 7 The three-phase current waveforms show that during the inertial operation period, the motor current is zero, while during the pulse injection phase, the three-phase current exhibits periodic current spikes, the contours of which contain information about the line back electromotive force. The three-phase current at the 0.2s input is very small and will not cause any impact on the system. From... Figure 8 The AB line voltage waveforms show that the line voltage is equal to the AB line back electromotive force during the inertial operation period of 0.1-0.12s. During the pulse injection phase of 0.12-0.2s, when the pulse is injected and the current continues, the absolute value of the line voltage is equal to the bus voltage. After the current continues, the line voltage is equal to the line back electromotive force.

[0119] Within 0.12–0.2 s, the single-branch current after pulse injection is acquired and the re-injection factor is calculated. , and Extraction, then according to Figure 5 The phase-locked loop structure shown is transformed to obtain the coordinate system in the stationary coordinate system. , Then normalization is performed to obtain , , Figure 9The normalized re-switching factor waveform obtained in this stage is shown. The two waveforms are 90° out of phase, resulting in a re-switching factor waveform with good sinusoidal performance in a very short time. Finally, the re-switching factor estimation result is fed into the subsequent phase-locked loop for rotor position and speed estimation.

[0120] Figure 10 The simulation results demonstrate the effectiveness of rotor position estimation during pulse injection. The estimated rotor position converges to the true angle within half an electrical angle cycle, exhibiting rapid convergence. Furthermore, the rotor position error remains near zero, indicating high accuracy. Even when the rotational speed dynamically changes during estimation, the rotor position estimation maintains high accuracy.

[0121] Figure 11 The simulation results demonstrate the effectiveness of speed estimation during pulse injection. The estimated speed converges to the true speed within a short time, exhibiting a fast convergence rate. Furthermore, the speed estimation error is within ±10 rpm. Even when the speed changes dynamically during estimation, the estimation maintains high accuracy.

[0122] Taking the single-branch current sampling topology and four-line voltage vector injection method as an example, in order to experimentally verify the rotor position and speed estimation effect of the proposed method at different speeds, the load machine first drags the controlled motor to a specified speed, and injects the line voltage vector into the inertial controlled motor system at a specific time according to the proposed method.

[0123] Figure 12 The proposed method is demonstrated to estimate the rotational speed and rotor position under inertial operation at 300 rpm (20% of rated speed). Figure 12 (a) shows the rotor position estimation effect of the proposed method. The left side is the overall estimation effect and the right side is a local magnified view. From the local magnified effect, it can be seen that the proposed method can achieve rotor position convergence within one electrical angle period. It has a large rotor position error only at the initial moment, and the rotor position error after steady state is kept within ±5°. Figure 12 (b) illustrates the speed estimation performance of the proposed method. The left side shows the overall estimation result, and the right side shows a magnified local view. The magnified view shows that the proposed method achieves speed convergence in a very short time, with a large speed estimation error only at the initial moment. The speed estimation error after steady state remains within ±20 rpm. In summary, Figure 11 The proposed method was verified to have a fast convergence speed under inertial operation at 20% of rated speed, and to have good rotor position and speed estimation effects. The steady-state estimation error is within the acceptable range for engineering applications.

[0124] Figure 13The proposed method is demonstrated to estimate the rotational speed and rotor position under inertial operation at 800 rpm (53% of rated speed). Figure 13 (a) shows the rotor position estimation effect of the proposed method. From the local magnification effect, it can be seen that the proposed method can still converge to the vicinity of the true rotor position within one electrical angle period. At the same time, it has a large rotor position error only at the initial moment, and the rotor position error after steady state is kept within ±3°. Figure 13 (b) demonstrates the speed estimation performance of the proposed method. As can be seen from the magnified view on the right, the proposed method achieves speed convergence in a very short time, with a large speed estimation error only at the initial moment. The speed estimation error after steady state remains within ±10 rpm. In summary, Figure 13 The proposed method was verified to have a fast convergence speed under inertial operation at 53% of rated speed, while also exhibiting good rotor position and speed estimation performance. Furthermore, compared to... Figure 12 The lower inertial operating speed, Figure 13 The steady-state estimation error is smaller compared to the previous method, and the error is also within the acceptable range for engineering.

[0125] Figure 14 The proposed method is demonstrated to estimate the rotational speed and rotor position under inertial operation at 1300 rpm (87% of rated speed). Figure 14 (a) shows the rotor position estimation effect of the proposed method. The left side is the overall estimation effect and the right side is a local magnified image. From the local magnified image, it can be seen that the proposed method converges to the vicinity of the true rotor position in about one electrical angle period. At the same time, it only has a large rotor position error at the initial moment, and the rotor position error after steady state is kept within ±3°. Figure 14 (b) demonstrates the speed estimation performance of the proposed method. As can be seen from the magnified view on the right, the proposed method achieves speed convergence in a very short time, with a large speed estimation error only at the initial moment. The speed estimation error after steady state remains within ±8 rpm. In summary, Figure 14 The proposed method was verified to have a fast convergence speed under inertial operation at 87% of rated speed, and also to have good rotor position and speed estimation results.

[0126] To verify the practical effectiveness of the proposed method, belt-speed restart experiments were conducted at three different inertial operating speeds (300 rpm, 800 rpm, and 1300 rpm). For each inertial operating speed, experiments were also conducted to increase or decrease the target speed by 200 rpm. Furthermore, for comparison, the effectiveness of direct restart during inertial operation was tested, and the differences between the proposed method and the direct restart method in terms of actual speed waveform, speed estimation error, rotor position error, and C-phase current amplitude were compared. It is worth noting that this invention simulates a wind turbine load, where the load torque increases quadratically with increasing speed. The sensorless scheme uses a static compensation voltage model observer, and the entire injection period is set to one carrier period. Additionally, the reference speed in this experiment exhibits a ramp-like change rather than a step.

[0127] Figure 15 The effect of belt-driven restart when the wind turbine is idled at 300 rpm under load is demonstrated, including the actual speed, speed and rotor position estimation error and C-phase current waveform under direct restart (left) and the proposed method (right). Figure 15 (a) shows a comparison of the restart effects at a target speed of 500 rpm (increased speed by 200 rpm after initial engagement). The experimental results show that the actual speed of the direct restart method on the left is forced to accelerate from zero speed to the target speed at the moment of engagement. The rotor position and speed estimation errors at the moment of engagement are relatively large, and the observer needs time to converge to the true value. At this time, the C-phase current shows a current spike at the moment of engagement, and then the current begins to increase from zero, driving the motor to accelerate. In contrast, the actual speed change of the method proposed on the right is smooth, without needing to accelerate from zero speed to the target speed again. Simultaneously, the speed and rotor position errors transition smoothly, and the C-phase current does not have excessively large spikes. (Figure) Figure 15 (b) The comparison of the restart effect at the target speed of 100 rpm (speed reduced by 200 rpm after activation) is shown. The experimental results show that the proposed method can achieve a smooth change in speed and a smooth transition from inertial operation to sensorless control compared with the direct restart method, and avoids large current spikes. Figure 15 (c) shows the C-phase current waveform during pulse injection, which is consistent with the theoretical analysis. When selecting the injection time, factors such as sampling coordination and the current drop rate of the specific system should be considered.

[0128] Figure 16 The effect of belt-driven restart when the fan is running at 800 rpm under load is demonstrated. Figure 16(a) shows a comparison of the restart effects at a target speed of 1000 rpm (with a 200 rpm ramp-up after initial engagement). The experimental results show that the actual speed of the direct restart method (left side) fluctuates drastically at the moment of engagement, and is forced to ramp up from zero speed to the target speed. The rotor position and speed estimation errors at the moment of engagement are large, and the observer needs time to converge to the true value. At this time, the C-phase current exhibits a large spike at the moment of engagement, and then the current begins to increase from zero, driving the motor to accelerate. In contrast, the actual speed change of the method proposed on the right side is stable, without needing to accelerate back to the target speed from zero speed. Simultaneously, the speed and rotor position errors transition smoothly, and the C-phase current does not exhibit excessive spikes. Figure 16 (b) The comparison of the restart effect at the target speed of 600 rpm (reduced by 200 rpm after activation) is shown. The experimental results show that the proposed method can achieve a smooth change in speed and a smooth transition from inertial operation to sensorless control compared to the direct restart method. It also avoids large phase current spikes and prevents unnecessary current protection or even damage to the system.

[0129] Figure 17 The effect of belt-driven restart when the fan is running at 1300 rpm under load is demonstrated. Figure 17 (a) shows a comparison of the restart effects at a target speed of 1500 rpm (with a 200 rpm ramp-up after initial engagement). The experimental results show that the actual speed of the direct restart method on the left fluctuates drastically at the moment of engagement, and is forced to ramp up from zero speed to the target speed. The rotor position and speed estimation errors at the moment of engagement are large, requiring time for the observer to converge to the true value. Furthermore, a large spike appears in the C-phase current at the moment of engagement, which could potentially trigger unnecessary current protection or even damage the system. In contrast, the actual speed change of the method proposed on the right is stable, requiring no re-acceleration from zero speed to the target speed. Simultaneously, the speed and rotor position errors transition smoothly, and the C-phase current does not exhibit excessive spikes. Figure 17 (b) The comparison of the restart effect at the target speed of 1100 rpm (speed reduction of 200 rpm after activation) is shown. The experimental results show that the proposed method can achieve a smooth change in speed and a smooth transition from inertial operation to sensorless control compared with the direct restart method, and avoids large phase current spikes.

[0130] In summary, the above experimental results, using a single-branch current sampling topology and a four-wire voltage vector injection method as examples, demonstrate the performance of the proposed method in low, medium, and high-speed domains for belt-speed re-switching, verifying the effectiveness, reliability, and superiority of the proposed belt-speed re-switching method.

[0131] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A single-current sensor-based speed-reconnection method based on line voltage vector injection, applicable to a permanent magnet synchronous motor system driven by a three-phase two-level inverter, characterized in that: The method includes: Define the switching state of the inverter, allowing only two phase arms to perform complementary switching actions, forcing all upper and lower transistors of the remaining phase arm to turn off, and generating a set of line voltage vectors including a single-phase floating phase. Based on the installation topology of the current sensor, vectors that can form an effective sampling path are selected from the line voltage vectors and used as the injection vector group; Within one injection cycle, vectors of the vector group are injected into the inverter sequentially, and before switching to the next vector, the motor winding current is allowed to decay to zero to ensure that the initial current value when each vector is applied is zero. Collect current data during the action of each injected line voltage vector, and combine the symmetry of motor winding parameters to establish an equivalent circuit equation that ignores stator resistance voltage drop; By performing pairwise operations on the equivalent circuit equations corresponding to different injected line voltage vectors and eliminating the bus voltage term, a linear relationship between the line back electromotive force and the rate of change of current is obtained. The re-start factor is defined as the difference in the current slope in the linear relationship between the line back electromotive force and the rate of change of current; the re-start factor corresponds to the line back electromotive force and includes the speed and rotor position information required for belt speed re-start. After normalizing the re-entry factor estimation results, the results are fed into the quadrature phase-locked loop. The output of the phase-locked loop is compensated based on the phase offset between the line back electromotive force and the rotor flux linkage to obtain the rotor position and speed information of the motor during inertial operation. The rotor position and speed information are transmitted to the sensorless control observer, the inverter is restarted, and the motor is seamlessly switched from inertial operation to sensorless control.

2. The single-current sensor belt-speed re-switching method based on line voltage vector injection according to claim 1, characterized in that: The single-phase floating line voltage vector includes 12 types of line voltage vectors, specifically: (00X), (01X), (10X), (11X), (0X0), (0X1), (1X0), (1X1), (X00), (X01), (X10), (X11), where "X" indicates that the upper and lower bridge arms of that phase are completely turned off.

3. The single-current sensor belt-speed re-switching method based on line voltage vector injection according to claim 1, characterized in that: When the current sensor is a single-branch current sampling topology, the injection vector group contains four vectors: (X10), (X01), (01X), and (0X1), which are defined as LV1(X10), LV2(X01), LV3(01X), and LV4(0X1), respectively.

4. The single-current sensor belt-speed re-switching method based on line voltage vector injection according to claim 3, characterized in that: The reinvestment factor is a linearly symmetric three-phase reinvestment factor. , , The specific expression is: In the formula, The electric angular velocity of the motor. It is a permanent magnet flux linkage. For rotor electrical angle, For phase inductance, For mutual induction, , , , These are the sampled values ​​of the motor phase current under the action of the corresponding line voltage vector; The three-phase re-entry factor is first subjected to Clark transformation and then fed into a quadrature phase-locked loop. The feedback angle inside the phase-locked loop needs to be superimposed with a phase compensation amount of π / 6.

5. The single-current sensor belt-speed re-switching method based on line voltage vector injection according to claim 4, characterized in that: The PI parameters of the orthogonal phase-locked loop are set according to the optimal damping ratio of the second-order system, satisfying: in, The proportional gain of the quadrature phase-locked loop. The integral gain of the quadrature phase-locked loop (PLL) is... It is the undamped natural frequency.

6. The single-current sensor belt-speed re-switching method based on line voltage vector injection according to claim 3, characterized in that: By optimizing the injection line voltage vector combination, the injection vector group is reduced to three vectors: LV1 (X10), LV2 (X01), and LV3 (01X). The reinvestment factor is orthogonally symmetrical. , The specific expression is: In the formula, The electric angular velocity of the motor. It is a permanent magnet flux linkage. For rotor electrical angle, For phase inductance, For mutual induction, , , These are the sampled values ​​of the motor phase current under the action of the injected vector group; The re-projection factor does not require Clark transformation and is directly fed into the quadrature phase-locked loop. The feedback angle inside the phase-locked loop needs to be superimposed with a phase compensation amount of π / 3.

7. The single-current sensor belt-speed re-switching method based on line voltage vector injection according to claim 1, characterized in that: When the current sensor is a DC bus current sampling topology, the injection vector group includes six vectors: (01X), (10X), (0X1), (1X0), (X01), and (X10), which are defined as LV1(01X), LV2(10X), LV3(0X1), LV4(1X0), LV5(X01), and LV6(X10), respectively.

8. The single-current sensor belt-speed re-switching method based on line voltage vector injection according to claim 7, characterized in that: By optimizing the injection vector combination, the injection vector group reduces the three vectors LV2 (10X), LV5 (X01), and LV6 (X10); The reinvestment factor is orthogonally symmetrical. , The specific expression is: In the formula, The electric angular velocity of the motor. It is a permanent magnet flux linkage. For rotor electrical angle, For phase inductance, For mutual induction, , , These are the sampled values ​​of the motor phase current under the action of the injected vector group; The re-projection factor does not require Clark transformation and is directly fed into the quadrature phase-locked loop. The feedback angle inside the phase-locked loop needs to be superimposed with a phase compensation amount of π / 3.

9. The single-current sensor belt-speed re-switching method based on line voltage vector injection according to claim 1, characterized in that: The current sampling time during each line voltage vector action period is set at the end of the vector injection cycle, with a preset sampling hold time reserved.

10. A single-current sensor-based speed-reset system based on line voltage vector injection, suitable for a permanent magnet synchronous motor system driven by a three-phase two-level inverter, characterized in that... The system includes: Vector generation module: used to define the switching state of the inverter, only allowing two phase arms to perform complementary switching actions, forcing all upper and lower transistors of the remaining phase arm to be turned off, and generating a line voltage vector including a single-phase floating phase. Vector filtering module: used to filter out vectors that can form an effective sampling path from the line voltage vectors based on the installation position of the single current sensor, and construct an injection vector group; Injection control module: used to sequentially inject the vectors in the injection vector group into the inverter within one injection cycle, and wait for the motor winding current to decay to zero before switching to the next vector, to ensure that the initial current value when each vector is applied is zero; Equation Establishment Module: Used to collect current data during the action of each injected line voltage vector, and combine the symmetry of motor winding parameters to establish an equivalent circuit equation that ignores the stator resistance voltage drop; Differential operation module: used to perform differential operation on the equivalent circuit equations corresponding to different injected line voltage vectors, eliminate the bus voltage term, and obtain the observation expression in which the line back electromotive force is linearly related to the rate of change of current; Re-start factor calculation module: used to define the re-start factor as the difference in current slope in the observation expression. The re-start factor is in phase with the line back electromotive force and includes the speed and rotor position information required for belt speed re-start. Phase-locked loop (PLL) calculation module: This module normalizes the re-entry factor and feeds it into the quadrature PLL. It compensates for the PLL output based on the inherent phase offset between the line back electromotive force and the rotor flux linkage, and calculates the rotor position and speed information of the motor during inertial operation. Observer initialization module: used to assign the rotor position and speed information to the initial state of the sensorless control observer, restart the inverter, and complete the seamless transition of the motor from inertial operation to sensorless control.

Images