A position sensorless control method for a two-phase switched reluctance motor

Abstract

The present invention discloses a position sensorless control method for a two-phase switched reluctance motor, which belongs to the technical field of motor control. Based on a special number of phases, the present invention adopts corresponding algorithms for different speed ranges, wherein a high-frequency pulse injection method with dynamic threshold update is adopted during medium and low speed operation, the algorithm sets a threshold at a specific position and dynamically adjusts it by calculating the speed to expand the estimation range of the traditional algorithm, and there is no need to partition the rotor position; under medium and high speed operation, a current gradient method for detecting multiple types of inductance inflection points is adopted, the algorithm collects phase current for identification and differential calculation, captures gradient transition points and sends out retrieval pulses, calculates the speed and recursively compensates to obtain the rotor position, the transition point usually occurs when the rotor is located at multiple types of inductance inflection points such as the beginning of the minimum inductance, the end of the minimum inductance and the beginning of the maximum inductance. The algorithms under different speed ranges are switched by setting the speed threshold and the corresponding flag bit, and finally realize the position sensorless control of the two-phase switched reluctance motor within the full speed range.

Description

Technical Field

[0001] The invention discloses a position sensorless control method applicable to the operation of a two-phase switched reluctance motor, and belongs to the technical field of motor control. Background Art

[0002] The switched reluctance motor (SRM) speed control system has the advantages of both DC speed control and AC speed control, and has a wide range of application prospects. As a type of the entire switched reluctance motor series, the two-phase SRM has more obvious advantages, including a further simplified structure, further reduced motor and controller costs, fewer connections, and larger slot space. Therefore, using two-phase SRM in a unidirectional high-speed speed control system can greatly reduce system complexity and control costs.

[0003] The key to ensure the efficient operation of SRM is that the motor winding completes the phase change at the precise rotor angle position. The position sensorless control strategy can replace the position sensor to improve the reliability of the system, reduce the complexity of the system, and accelerate the practical application of SRM in various fields. However, the current research on the position sensorless control technology of SRM is mainly focused on three-phase and other conventional phase structures. In this type of structure, high-frequency pulses are usually injected into the non-conducting phase, and the pulse current signals of multiple non-conducting phases are sampled for logical operation to determine the real-time position of the rotor. In the operation process of the two-phase SRM, only the current signal of a single non-conducting phase is available for sampling and analysis, so the existing position sensorless control method cannot be directly applied to the two-phase structure. In view of this, the present invention proposes a position sensorless control algorithm suitable for two-phase SRM. The method adopts a high-frequency pulse injection method with real-time threshold update in the medium and low speed operation stage, combines multiple types of inductance inflection points to detect the current gradient in the medium and high speed stage, and realizes the full-speed domain position sensorless control of the two-phase SRM by setting the speed threshold and the corresponding flag. Summary of the invention

[0004] The present invention proposes a two-phase switched reluctance motor position sensorless control method for a two-phase SRM. When the motor is in the medium and low speed operation stage, the method adopts a high-frequency pulse injection method with a dynamically updated threshold to predict the rotor position, without partitioning the rotor position and weakening the adverse effect of the motor motion back electromotive force on the control accuracy; in the medium and high speed stage, the winding current gradient jump caused by various inductance inflection points is detected and the motor speed and rotor position are calculated based on it, which can adapt to various current waveforms in the operation of the motor, and the control is convenient and flexible. The control method proposed in the present invention can be applied to the SRM of a two-phase structure to realize full-speed operation, and smooth switching between low-speed and high-speed algorithms can be realized by setting the speed threshold and the corresponding flag bit. At the same time, the control strategy does not need to establish an accurate mathematical model of the motor or add additional hardware equipment, and can further improve the reliability of the drive system on the basis of a small number of phases and low cost. In order to achieve the above technical objectives, the present invention adopts the following technical solutions to achieve it.

[0005] A position sensorless control method for a two-phase switched reluctance motor comprises the following steps:

[0006] 1. The switched reluctance motor is driven by single-phase alternating conduction. The conduction angle of each phase in the two-phase structure is approximately π;

[0007] 2. Set a speed threshold. When the speed is lower than the threshold, a high-frequency pulse injection method with dynamic current threshold adjustment is used as shown in step 3. Otherwise, a current gradient method based on multiple types of position points is used as shown in step 4.

[0008] 3. High-frequency pulse injection method to overcome the influence of back electromotive force:

[0009] (1) Setting a current threshold, selecting a fixed frequency and duty cycle when the off-phase current is less than the threshold, and injecting high-frequency pulses into the corresponding phase using voltage PWM;

[0010] (2) Collect the pulse current peak value at the falling edge of the pulse injection control signal and keep it until the next sampling time;

[0011] (3) The electromagnetic characteristics of the motor are obtained through finite element simulation or actual measurement and the initial threshold is calculated using the minimum inductance;

[0012] (4) comparing the peak value of the pulse current with the threshold value, and sending a position search pulse when the peak value is not less than the threshold value;

[0013] (5) calculating the real-time speed of the motor and the corresponding rotor angle through adjacent search pulses, and transferring the calculated speed into the threshold calculation model to obtain the threshold value at the corresponding speed;

[0014] (6) Feedback the threshold corresponding to the real-time speed to d and repeat the above process;

[0015] 4. Current gradient method based on multiple types of inductance inflection points:

[0016] (1) Detecting the phase current and performing differential calculation to obtain the phase current gradient;

[0017] (2) Detect the special moment when any type of phase current gradient jumps: 1. The jump point from positive to negative; 2. The jump point from negative to positive and send out a search pulse;

[0018] (3) Calculate the real-time speed of the motor and the corresponding rotor angle position through adjacent retrieval pulses;

[0019] 5. The rotor position angle estimated by the algorithm replaces the actual position to control the motor operation.

[0020] The single-phase alternating conduction method adopted by the two-phase switched reluctance motor is as follows: in the low-speed conduction stage, each phase winding is subjected to hysteresis chopping control in the conduction interval of the phase, and the chopping limit is determined by the output of the difference between the predetermined speed and the real-time speed after PI adjustment. The loop width is obtained according to the switching tube frequency and the bus voltage, and high-frequency short-time pulses are injected into the non-conduction interval. Among them, the opening angle is fixed at the minimum inductance in the low-speed stage, and the turn-off angle is adjusted considering the freewheeling time so that the conduction angle is π. In the high-speed conduction stage, the opening angle and the turn-off angle are advanced, and the conduction angle is appropriately reduced.

[0021] The two-phase switched reluctance motor position sensorless control method includes a high-frequency pulse injection method with a single threshold dynamic adjustment under low-speed operation and a current gradient method based on multiple types of inductance inflection points under high-speed operation. Furthermore, the algorithm sets a speed threshold for real-time comparison and performs algorithm switching.

[0022] The initial threshold of the high-frequency pulse injection method is Calculated, where U k is the phase winding voltage, L k is the phase inductance, f pulse is the pulse injection frequency, and D is the pulse injection duty cycle. Further, the present invention calculates the initial threshold and the subsequent dynamic threshold based on the position of the minimum inductance.

[0023] The high-frequency pulse injection method collects the peak value of the pulse current and compares it with the dynamically changing threshold value to send out a retrieval pulse, and calculates the motor speed and rotor position angle through adjacent pulse intervals. Furthermore, the algorithm does not need to partition the rotor position and overcomes the influence of motion back electromotive force, thereby improving the estimation range of the algorithm.

[0024] The high-frequency pulse injection method can calculate the corresponding threshold value at any position and send out a search pulse. In the present invention, the threshold value is calculated and compared at the minimum inductance position. This position is the opening angle position when running at low speed. As the starting position, the compensation link can be omitted to reduce the error, and this position is not affected by the inductance saturation characteristics. Further, the high-frequency pulse injection method can realize variable angle control, and any position angle can be used as the opening angle of the phase winding to compare with the threshold value through mathematical fitting of the inductance curve model.

[0025] The current gradient method collects phase current for differential calculation and captures the jump points of the current gradient, which include points from positive to negative and points from negative to positive. Furthermore, when changing from positive to negative, the rotor position is at the end of the minimum inductance or the beginning of the minimum inductance under zero-voltage freewheeling, and when changing from negative to positive, the rotor position is at the beginning of the maximum inductance value.

[0026] The current gradient method sends out a search pulse at the position where the current gradient jumps. The controller calculates the real-time speed of the motor based on the edge moments of two adjacent pulses, obtains the relative position angle at a special position through finite element calculation or numerical simulation method, and finally compensates the position angle obtained by linear recursion of the speed to obtain the real-time rotor position angle.

[0027] The system for implementing the two-phase switched reluctance motor position sensorless control method includes: a current sensor and a transmission module for obtaining phase current information of two-phase windings, a controller for generating bridge drive signals and speed and position angle information of power converters of each phase winding according to the phase current information, and an asymmetric half-bridge power converter and a drive circuit for motor driving.

[0028] The present invention adopts the above technical solution and has the following technical effects: the present invention discloses a two-phase switched reluctance motor position sensorless control method, which uses a high-frequency pulse injection method with dynamically updated thresholds to calculate the speed and rotor position at low speeds, and uses a current gradient method that detects multiple types of inflection points of inductance to linearly recursively calculate the rotor position at high speeds. At the same time, the two algorithms are switched by setting the speed threshold and the flag bit to achieve continuous rotor position estimation over the full speed range. This method does not require any prior knowledge of the motor or additional hardware circuits, overcomes the defect that the traditional low-speed control algorithm is limited by the motion back electromotive force, has a simple implementation process, high positioning accuracy and is not affected by electromagnetic saturation characteristics, and can be applied to two-phase switched reluctance motors to further improve the reliability of the drive system and reduce maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0030] Figure 1 The schematic diagram of the control object in the present invention - the main body and driving circuit of the two-phase 8 / 4-pole switched reluctance motor;

[0031] Figure 2 Schematic diagrams of different working modes of the two-phase switched reluctance motor drive circuit in the present invention, (a), (b), and (c) respectively represent conduction, zero-voltage freewheeling, and reverse-voltage freewheeling;

[0032] Figure 3 This is a schematic diagram of the principle of high-frequency pulse injection;

[0033] Figure 4 A schematic diagram of the current waveform of a two-phase switched reluctance motor using current chopping control in the conduction interval and high-frequency pulse injection in the non-conduction interval;

[0034] Figure 5 It is a schematic diagram of the principle of the high-frequency pulse injection method of the present invention with dynamic change of current threshold under low-speed operation;

[0035] FIG6( a ) is a schematic diagram of the principle of the current gradient method based on the minimum inductance end under high-speed operation in the present invention;

[0036] FIG6( b ) is a schematic diagram of the principle of the current gradient method based on the maximum inductance starting point under high-speed operation in the present invention;

[0037] FIG6( c ) is a schematic diagram of the principle of the current gradient method based on the minimum inductance starting point under high-speed operation in the present invention;

[0038] Figure 7 The algorithm flow chart of the full-speed range position sensorless control method of the two-phase switched reluctance motor;

[0039] Figure 8 This is a curve comparison diagram of the high-frequency pulse injection method with dynamic threshold update and the current gradient method for detecting the inflection points of multiple types of inductors.

[0040] in, Figure 1 , 2 A 1-4 Indicates the stator poles wound by the same phase winding, Q 1、2 Indicates the power switch tube, D 1、2 represents diode, U represents bus voltage; Figure 3 In , Δt represents the pulse injection time; Figure 4 In the equation, i and L represent phase current and phase inductance respectively, θ on ,θ off They represent the on and off angles of the windings respectively; Figure 5 In, L max , L minThey respectively represent the maximum inductance and minimum inductance of the phase winding, θ1 and θ2 are the rotor positions when the inductance reaches the minimum value twice in any electrical cycle under low-speed operation; in Figures 6-8, θ2, θ3, θ4, and θ5 respectively represent the rotor positions at: the end of minimum inductance, the beginning of maximum inductance, the end of maximum inductance, and the beginning of minimum inductance. DETAILED DESCRIPTION

[0041] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0042] The control object in the control algorithm proposed by the present invention is a two-phase switched reluctance motor. Figure 1 Take the two-phase 8 / 4-pole switched reluctance motor shown in as an example, any one-phase winding corresponds to 4 of the 8 stator poles. Taking phase A as an example, the winding is wound around poles A1-A4 in sequence every other stator pole, and the four sets of windings on poles A1-A4 can be connected in series or in parallel. Figure 1 The windings belonging to the A phase are connected in series in sequence. The external power supply is connected to the motor winding through a power converter, and the beginning of the winding wound on A1 and the end of the winding wound on A4 are connected to the positive and negative ends of the power supply voltage through switch tubes Q1 and Q2 respectively. D1 and D2 are freewheeling diodes when the winding is turned off, among which D1's cathode is connected to the positive end of the power supply voltage and its anode is connected to the end of the A phase winding, and D2's anode is connected to the negative end of the power supply voltage and its cathode is connected to the beginning of the A phase winding. The physical parameters of the two-phase windings are exactly the same, and they differ only by a mechanical angle of π / 4 in spatial arrangement. When the motor is operating normally, the corresponding drive circuit has three different working states, as follows: Figure 2 As shown, taking the A-phase winding as an example, when the current passing through the winding is not zero, (a), (b), and (c) respectively indicate that the A-phase winding is in the normal conduction state, the zero-voltage freewheeling state, and the reverse voltage freewheeling state. When in the normal conduction state, Q1 and Q2 are turned on, and the power supply voltage supplies power to the A-phase winding through two switching tubes. The current loop and direction are shown in the blue and red loops in (a); when in the zero-voltage freewheeling state, only one switching tube is turned on and the winding phase voltage is zero. Taking Q1 as an example, at this time, the A-phase winding is freewheeling with Q1 through the diode D2, and the current loop and direction are shown in the blue and red loops in (b); when in the reverse voltage freewheeling state, both switching tubes are turned off, and the A-phase winding returns to the power supply through the diodes D1 and D2. The current loop and direction are shown in the blue and red loops in (c).

[0043] For the above control object, the control method proposed by the present invention at low speed is to inject high-frequency pulses into the non-conducting phase. The principle is as follows: Figure 3 As shown in . During the pulse injection period, Q1 and Q2 are turned on at the same time, and the voltage equation of the kth phase of the switched reluctance motor is:

[0044]

[0045] In formula (1), U k is the kth (k=1,2) phase winding terminal voltage, i k is the winding current of phase k, R k , ψ k , L k are the winding resistance, flux linkage, and inductance of the kth phase, respectively, and θ is the rotor position. When the motor is stationary and running at a low speed after starting, because the frequency of the injected pulse is very high, the response current amplitude is very small, and the influence of factors such as the motor's motion back electromotive force, electromagnetic saturation effect, and winding voltage drop can be ignored. The response current after the pulse is injected is expressed by formula (2):

[0046]

[0047] Where: i pk is the response current amplitude, L(θ) is the inductance of the motor, U dc is the bus voltage, Δt is the pulse injection time, f pulse is the pulse injection frequency, and D is the pulse injection duty cycle. From formula (2), it can be seen that when the bus voltage and the detection pulse frequency are given, the pulse response current amplitude of the non-conducting phase is inversely proportional to the phase inductance. Figure 2 (c) The reverse voltage freewheeling mode is to turn off the upper and lower bridge arm switches at the same time, so the pulse current rising and falling processes are symmetrical.

[0048] The two-phase switched reluctance motor in the present invention adopts a single-phase alternating conduction method at low speed. The specific implementation method is: 1. The turn-on angle is fixed at the minimum inductance, and the turn-off angle is appropriately advanced considering the freewheeling time so that the conduction angle is π / 4, that is, the electrical angle is π; 2. The current chopping control method is adopted in the conduction interval, and the chopping limit is determined by the output of the difference between the predetermined speed and the real-time speed after PI adjustment. The loop width can be flexibly adjusted according to the maximum allowable frequency of the switch tube and the bus voltage; 3. After the phase current is detected to be less than a certain threshold in the non-conduction interval, a high-frequency pulse is injected. After the above measures are implemented, the current waveform of the two-phase winding during the low-speed operation of the motor is as follows Figure 4 As shown, it can be seen that the high-frequency pulse is only injected into the non-conducting phase, so the pulse is injected into the corresponding phase only when the inductance change rate is negative or the rotor pole is facing the stator slot. From the figure, the pulsation curve of the corresponding current peak can be obtained. The change trend of this curve is opposite to the trend of the inductance curve of the corresponding phase and the change pace is consistent, that is, the pulse response current peak of the non-conducting phase is inversely proportional to the phase inductance. However, as the speed increases, the motor's motion back electromotive force will gradually increase, so the pulse current amplitude and the threshold at a fixed position are decreasing with the speed, which is not considered in formula (2).

[0049] In view of the above situation, the present invention proposes a high-frequency pulse injection method with a variable threshold. The specific principle is as follows: Figure 5 Taking the two-phase motor as an example, θ1 and θ2 in the figure are the rotor positions where the inductance reaches the minimum value twice in the operation of the motor. At this time, all the electrical parameters of the winding are completely consistent, and the voltage equation is shown in formula (1). When the winding voltage drop is ignored, it can be obtained

[0050]

[0051] At this time, a fixed frequency voltage PWM control method is used to inject high-frequency pulses, and the duty cycle is fixed, that is, the pulse injection time is a constant. Since the rotor position is fixed during pulse injection, the inductance change rate at θ1 and θ2 is also a constant. At this time, the current peak values ​​at the two points can be obtained as:

[0052]

[0053] Where C1 = dL k / dθ| θ＝θ1 , C2=L min / Δt, when ω=0, i max =U k Δt / L min , L min The electromagnetic characteristics of the motor can be estimated by finite element calculation, so the current peak value when ω = 0 can be directly calculated. The relationship between the current peak value and the speed at the minimum inductance can be obtained through equation (4). If the peak value is set as the initial threshold, t Ax represents the xth time when the pulse current peak reaches the threshold, Δt_x represents the difference between the xth time when phase A reaches the threshold and the previous time when it reaches the threshold, and the motor speed can be obtained as:

[0054]

[0055] Among them, N r Indicates the number of rotors. In this embodiment, N r =4. After calculating the real-time motor speed, observe the current peak value i obtained at the minimum inductance position at this time. max , and substitute it into formula (4) to get C1. Therefore, the current threshold at any time when the motor is running at this special position point can be obtained, and the rotor position angle at any time can be calculated according to formula (6):

[0056]

[0057] When the rotor is near the minimum inductance position, the inductance is almost unaffected by the electromagnetic saturation characteristics as the current increases; using this position as the opening angle can quickly establish the current in the initial excitation stage, which is beneficial to improving the torque output capacity; and this position is used as the starting position in the angle position calculation model without the need for table lookup for compensation calculation, which is beneficial to reducing the estimation error. It should be noted that this method is applicable but not limited to using the minimum inductance position as the comparison moment for the opening and threshold calculation, and is only used as an example. When variable angle control is required, the inductance curve can be mathematically fitted to obtain the inductance value corresponding to any position, so as to obtain the value of the inductance corresponding to the position. Figure 5 As shown in the figure, the simulated inductance characteristic is a sine curve and the minimum value is L min According to the sinusoidal curve characteristics, it is easy to get the inductance value L at any position. k As the comparison moment between the opening angle and the threshold value, then L k Replace L min Repeat the above steps cf to estimate the angular position.

[0058] When the two-phase switched reluctance motor is in the high-speed operation stage, in order to maintain the same accuracy in the same non-conduction interval, the pulse injection frequency must be increased synchronously with the speed. However, due to the switching loss and the skin effect under high-frequency conditions, the injection frequency should not be too high. Therefore, the fixed pulse injection frequency will cause the number of pulses to decrease as the speed increases, which will inevitably reduce the position estimation accuracy. In view of this, the present invention adopts a current gradient method that captures the current gradient transition point and calculates the rotor position. The transition points captured by this method include two types: points from positive to negative and points from negative to positive. In the specific embodiment, three inductance inflection points are mainly listed to illustrate the algorithm in turn: the end of the minimum inductance value, the beginning of the maximum inductance value, and the beginning of the minimum inductance value. Among them, the end and beginning of the minimum inductance value belong to the first type of transition point, and the beginning of the maximum inductance value belongs to the second type of transition point.

[0059] In the inductance model, assuming that the main magnetic line is linear and the flux edge effect is not considered, the change of inductance with respect to the rotor position is linear as shown in Figure 6-8. In the linear inductance model, the winding phase voltage equation shown in equation (1) can be simplified to:

[0060]

[0061] When the motor is in the high-speed operation stage, the high back electromotive force of the winding leads to a low current building capability. Therefore, the turn-on angle should be appropriately advanced to increase the conduction current and improve the torque output capability in the inductance rising range. Accordingly, the turn-off angle should be appropriately advanced to avoid tail current. on ,θ2] interval, L=L min , the inductance change rate is 0, and the phase current change rate is:

[0062]

[0063] It can be concluded that in the minimum phase inductance interval, the phase current change rate is a fixed positive value, and the phase current rises linearly in this interval. In the phase inductance rising area, the phase inductance increases linearly, and the phase inductance at any angle θ in this interval can be expressed as L = L min +K(θ-θ2), where K is the inductance rising slope, so the phase current change rate in this interval is:

[0064]

[0065] When the opening angle θ on Ahead and meet θ on <θ2-L min / K, according to formula (9), it can be inferred that the phase current change rate is always negative in the linear rising area of ​​inductance. Combined with the phase current change rate in the minimum inductance interval, it can be known that the phase current change rate will change from positive to negative at the initial overlap position of the stator and rotor. Therefore, the position can be captured by detecting the jump point and sending a search pulse. The implementation principle is shown in Figure 6(a). After sending the search pulse, the real-time speed of the motor can also be calculated by formula (5). At this time, t Ax represents the xth jump time of the current gradient of phase A, and Δt_x represents the difference between the xth and the previous time when phase A reaches the minimum inductance end position. Similarly, by using equation (6) for phase compensation, the rotor position at any time under the current gradient method based on the minimum inductance end can be obtained:

[0066]

[0067] Where θ2 is the position angle at the end of the minimum inductance interval. This value can be obtained through the design parameters of the motor. In actual operation, it can also be obtained through finite element calculation or position error compensation.

[0068] Furthermore, in the interval [θ3, θ4], L = L max According to the linear inductance model, the phase current change rate at this time is:

[0069]

[0070] That is, the inductance change rate is positive in the interval where the inductance is maintained at the maximum value. According to formula (9), when the opening angle satisfies θ on <θ2-L min / K, the phase current change rate is negative in the inductance rising area. Combining with formula (11), it can be concluded that at the beginning of the maximum inductance value, that is, when θ=θ3, the phase current change rate will change from negative to positive. Capturing this transition point can send out a search pulse. The principle is shown in Figure 6(b). At this time, the speed can be calculated according to formula (5), t Axrepresents the xth jump moment of the current gradient of phase A, Δt_x represents the difference between the xth and previous moments when phase A reaches the starting position of the maximum inductance. Similarly, θ3 can be replaced by θ2 to calculate the rotor position angle at any moment through formula (10).

[0071] Accordingly, when θ2 is taken as the starting angle, assuming that the current after the relevant disconnection can continue to flow until the end of the electrical cycle under high-speed operation, and the winding is in a zero-voltage continuous current state in the interval [θ4, θ5], that is, in the inductance reduction area, as shown in Figure 2 (b) shows that L = L max -K(θ-θ4), K represents the inductance decreasing slope and its value is equal to the inductance increasing slope. Substituting it into the voltage equation (1), it can be seen that the phase current change rate in this interval is:

[0072]

[0073] Similarly, when the winding is in the zero-voltage freewheeling state and in the interval [θ5, π / 2], L = L min , the phase current change rate is:

[0074]

[0075] Therefore, when the winding is in the zero-voltage freewheeling state and the motor speed ω>R / K, the phase current change rate in the inductance drop zone is positive. Since the phase current change rate in formula (13) is negative, it can be concluded that when θ=θ5, the phase current change rate will change from positive to negative. Similarly, the transition point can be captured to send a retrieval pulse. The implementation principle is shown in Figure 6(c). According to formula (5), the speed is calculated, t Ax represents the x-th jump moment of the current gradient of phase A, Δt_x represents the difference between the x-th and the previous time when phase A reaches the starting position of the minimum inductance. Similarly, θ5 can be replaced by θ2 to calculate the rotor position angle at any time through formula (10).

[0076] In terms of algorithm switching, the present invention adopts a method of setting a speed flag. Since both the low-speed and high-speed algorithms can calculate the real-time speed, the speed flag can be obtained by comparing it with the threshold set in advance through feedback. The flag value is the difference between the threshold and the real-time speed, and its sign can determine the algorithm to be used at the corresponding speed. When the speed flag is positive, the low-speed algorithm is used, and vice versa. The setting of the speed threshold is related to the estimation range of the low-speed and high-speed algorithms. Generally, the threshold should be set in the intersection area within the estimation range of the two types of algorithms, that is, within the range set in this area, both types of algorithms can correctly calculate the rotor position.

[0077] In summary, the algorithm process disclosed in the present invention is as follows Figure 7As shown, at the beginning of each electrical cycle, the phase current is detected, and the speed is calculated by formula (5) to see whether it reaches the threshold. The speed flag is obtained by subtraction and its sign is determined. If it is positive, the high-frequency pulse injection method with dynamic threshold change mentioned above is used, and the real-time speed is substituted into formula (4) to calculate the current peak threshold and the rotor position is calculated by formula (6) until the end of the current electrical cycle; if it is negative, any type of phase current gradient method for detecting inductance inflection point can be used according to the current waveform, and the real-time speed is substituted into formula (10) to calculate the rotor position until the end of the current electrical cycle. At the beginning of each cycle, both algorithms will linearly recursively calculate the speed to obtain the real-time rotor position until the end of the current electrical cycle, and continuously update the system state and cycle in each electrical cycle to calculate the rotor position in the full speed range. The curve comparison of the high-frequency pulse injection method with dynamic threshold update and the current gradient method for detecting multiple types of inductance inflection points is shown in the figure. Figure 8 shown.

Claims

1. A position sensorless control method for a two-phase switched reluctance motor, characterized in that: The steps include: (1) Real-time detection and acquisition of phase current signals, and calculation of real-time rotation speed based on adjacent characteristic position points; (2) Setting a speed threshold and subtracting it from the calculated speed to obtain a speed flag. When the flag is positive, the rotor position is calculated using the method shown in step (3). Otherwise, the rotor position is calculated using the method in step (4). (3) High-frequency pulse injection method with threshold dynamically updated with real-time speed: Inject high-frequency short-time pulses with fixed frequency and duty cycle during the non-conduction period of the two-phase windings, set the initial threshold based on any rotor position point, and send a position retrieval pulse when the peak value of the pulse current is not lower than the threshold. Finally, calculate the real-time speed and rotor position through the interval between adjacent pulses and update the threshold state. (4) Phase current gradient method for detecting multiple types of inductance inflection points: Based on multiple types of phase current waveforms, multiple types of inductance inflection points are detected, that is, the moment when the phase current gradient jumps and a search pulse is issued, and then the real-time speed and rotor position are calculated through the interval between adjacent pulses. Among them, the jump point includes two types: from positive to negative and from negative to positive, and phase compensation needs to be performed according to different position points; (5) The rotor position angle estimated by the algorithm replaces the actual position to control the motor operation; The initial threshold is calculated by phase voltage, inductance, and pulse injection time, and the calculation formula is: ， in, represents the initial threshold, , , , They represent phase voltage, voltage chopping duty cycle, minimum inductance and pulse injection frequency respectively. The dynamic threshold is obtained by establishing a mathematical model of speed and voltage and compared with real-time feedback, and a search pulse is sent out at a time not less than the threshold. The specific relationship is: ， in, Indicates the real-time threshold, ω Indicates the speed calculated in real time. C 1 and C 2 are all constants.

2. A position sensorless control method for a two-phase switched reluctance motor according to claim 1, characterized in that: It includes a high-frequency pulse injection method for dynamically updating a single threshold value under low-speed operation and a current gradient method for detecting multiple types of inductance turning points under high-speed operation. The high-frequency pulse injection method and the phase current gradient method are compared and switched in real time by setting a speed threshold.

3. A position sensorless control method for a two-phase switched reluctance motor according to claim 2, characterized in that: The high-frequency pulse injection method can calculate the corresponding threshold at any position and send out a search pulse; it can realize variable angle control, and through the mathematical fitting of the inductance curve model, it can realize any position angle as the opening angle of the phase winding and the threshold comparison moment. When the threshold is set at the minimum inductance position, no phase compensation is required and the influence of the inductance saturation effect can be eliminated.

4. A position sensorless control method for a two-phase switched reluctance motor according to claim 2, characterized in that: In the current gradient method, phase current signals are collected, and different types of inductance inflection points are captured according to different current waveforms. Two types of jump points are specifically manifested on the current gradient, including points from positive to negative and points from negative to positive. When changing from positive to negative, the rotor position is at the end of minimum inductance or the beginning of minimum inductance under zero-voltage freewheeling conditions, and when changing from negative to positive, the rotor position is at the beginning of maximum inductance.

5. A position sensorless control method for a two-phase switched reluctance motor according to claim 2, characterized in that: In the current gradient method, a search pulse is emitted at the position where the current gradient jumps and the speed is calculated accordingly. The relative position angle at a special position is obtained through finite element calculation or numerical simulation method. Finally, the position angle obtained by linear recursion of the speed is compensated to obtain the real-time rotor position angle.

6. A control system for implementing a position sensorless control method for a two-phase switched reluctance motor as claimed in any one of claims 1 to 5, characterized in that: include: A current sensor and a transmission module for obtaining winding phase current information; a controller for generating a two-phase winding power converter bridge drive signal and speed and position angle information according to the phase current information; The invention discloses an asymmetric half-bridge power converter and a driving circuit for motor driving.

Images