JP6056629B2 - Motor control device - Google Patents

Description

  The present invention relates to a motor control device that PWM-controls a brushless DC motor by a position sensorless system.

  As a driving method of the brushless DC motor, there is a so-called position sensorless method in which a zero cross point of an induced voltage generated in a stator winding of the motor is detected and the energized phases are sequentially switched with reference to the zero cross point. At this time, if the motor is PWM-controlled, the pulse of the PWM signal is superimposed on the induced voltage as noise, so in order to accurately grasp the zero cross point, how to eliminate the influence of the noise becomes a problem. .

  For example, Patent Document 1 discloses a technique for estimating the timing at which the induced voltage reaches a neutral point voltage (zero cross point) using waveform data corresponding to the number of rotations of the motor. In addition, there are techniques such as giving up the detection of zero-cross points at the timing when noise is generated and removing noise by filtering.

Japanese Patent Laid-Open No. 9-9676

  However, the prior art as described above is not suitable for a system that requires high-speed responsiveness because a response delay due to filter processing occurs. In addition, since an interrupt is generated in the CPU to switch the energization pattern of each phase, a high processing capability is required for the CPU to rotate the motor at high speed. However, when the load is occupied by the filter processing for detecting the zero cross point, the CPU cannot perform other controls, and it becomes difficult to integrate many functions into the control circuit including the CPU.

  The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a motor control device that can acquire the energization timing based on the zero cross point of the induced voltage without increasing the processing load of the CPU.

According to the motor control device of the first aspect, the control circuit having a CPU estimates the zero-cross point of the induced voltage based on the terminal voltage of the stator winding of the brushless DC motor, and transfers it to the stator winding. The PWM command value calculated based on the rotational speed of the motor measured from the estimated zero-cross point interval is output. Specifically, the control circuit sets the first detection timing for the terminal voltage of the stator winding that becomes the non-conduction phase after a predetermined time has elapsed from the start of the PWM cycle, and then after the predetermined time has elapsed. A second detection timing of the terminal voltage is set.

  Then, if one of the terminal voltages detected at the first and second detection timings is lower than the neutral point voltage of the motor and the other of the terminal voltages exceeds the neutral point voltage, the first and second detection conditions are satisfied. The middle of the second detection timing is estimated as the zero cross point, and the first and second detection timings in the next detection cycle are determined based on the estimated zero cross point. On the other hand, if the detection condition is not satisfied, the time between the first and second detection timings in the next PWM cycle is set to be longer.

  That is, if the window between the first and second detection timings is a window for capturing the zero-cross point, it is considered that the zero-cross point is in the window when the detection condition is satisfied. Therefore, it is possible to reasonably estimate the middle of the window time width as the zero cross point without directly detecting the zero cross point. On the other hand, if the detection condition is not satisfied, the zero cross point is considered to be outside the window. Therefore, the time width of the window used for the next detection cycle is made longer so that the zero cross point can be captured within the window.

  As a result, the CPU control program (software) constituting the control circuit can set the energization timing based on the zero-cross point of the induced voltage without performing the filter process, so that the CPU processing load is reduced compared to the conventional case, and other than motor control. These functions can also be executed in parallel.

Functional block diagram showing the configuration of the motor drive control device according to the first embodiment The figure which shows the outline | summary of control about the three-phase induced voltage waveform of a motor Flow chart showing control contents (A), (b) is a flowchart showing the processing contents of steps S2 and S4 in FIG. 3, (c) is a diagram showing detection (1) and (2) phase corresponding to each energization pattern. Flow chart showing processing details of free-run determination Diagram explaining switching of energization pattern A diagram conceptually explaining energization pattern switching control by a microcomputer Diagram showing motor startup sequence Assuming that the microcomputer directly detects the zero cross point, (a) shows the case where the motor is rotating at a constant speed, and (b) shows the case where the motor is accelerating. FIG. 9 equivalent diagram showing the case where the motor is decelerating FIG. 9 equivalent diagram in the case of this embodiment Fig. 10 equivalent The flowchart which is 2nd Embodiment and shows learning control of energization time. A diagram explaining the content of learning control, with the horizontal axis representing the actual number of revolutions and the vertical axis representing the energization time.

(First embodiment)
As shown in FIG. 1, for example, an in-vehicle motor drive control device 1 is configured around a microcomputer (a microcomputer, a control circuit) 2 and an inverter circuit 3. The motor drive control device 1 is supplied with drive power via an ignition switch (IG) of the vehicle. The driving power is supplied to the inverter circuit 3 (driving circuit) via a π-type filter 6 including capacitors 4 a and 4 b and a coil 5.

  The driving power is supplied to the 5V power circuit 8 through the diode 7. The 5V power supply circuit 8 steps down the drive power supply voltage to generate a 5V control power supply, and supplies it to the power supply terminal of the microcomputer 2. A bypass capacitor 9 is connected between the power supply terminal of the microcomputer 2 and the ground. A motor speed command is input to the input circuit 10 as a low-speed PWM signal from a host controller (not shown). The input circuit 10 generates a higher-speed PWM signal corresponding to the input PWM command value (PWM duty) and outputs it to the input port of the microcomputer 2.

  The inverter circuit 3 includes three element modules 11U, 11V, and 11W, and each element module 11 includes two N-channel MOSFETs (switching elements) 12 and 13 connected in series. The common connection point of these FETs 12 and 13 becomes each phase output terminal of the inverter circuit 3, and is connected to each phase stator winding 15U, 15V, 15W of the motor 14 which is a three-phase brushless DC motor, for example. The microcomputer 2 (PWM signal output means) outputs a PWM signal to the gates of the N-channel MOSFETs 12 and 13 constituting the inverter circuit 3 by operating the built-in CPU in accordance with a control program (software). Drive control.

  The element module 11 also includes a comparator 16 (only the U phase is shown in FIG. 1) that compares the potential at the common connection point of the FETs 12 and 13 with the neutral point potential of the motor 14 (the sum of the voltages of each phase). Yes. The comparator 16 is used by the microcomputer 2 to estimate the zero cross point of the induced voltage of the motor 14. The output signal of the comparator 16 provided in each element module 11 is input to the input port of the microcomputer 2.

  A series circuit of a resistance element 17 and a thermistor 18 is connected between the output terminal of the 5V power supply circuit 8 and the ground. The thermistor 18 (only U phase is shown in FIG. 1) is disposed in the vicinity of each element module 11 and used to detect the temperature of the N-channel MOSFETs 12 and 13. A common connection point of the resistance element 17 and the thermistor 18 (temperature detection means) is connected to the A / D conversion input port of the microcomputer 2.

  The shunt resistor 19 is inserted in the ground line between connection points to which one ends of the capacitors 4a and 4b are connected. The terminal voltage of the shunt resistor 19 (current detection means) is input to the A / D conversion input port of the microcomputer 2 when amplified by the amplifier 20 (current detection means). Further, a series circuit (voltage detection means) of the resistance elements 21 and 22 is connected between the power supply terminal IG and the ground, and the common connection point thereof is connected to the A / D conversion input port of the microcomputer 2. Has been.

  In the present embodiment, the microcomputer 2 refers to the output signal of the comparator 16 (U, V, W) given to the input port, so that the stator windings 15U, 15V, 15W when the motor 14 is rotating. The zero crossing point of the induced voltage (a point crossing the neutral point voltage level) is estimated. Hereinafter, the principle will be described.

  Each phase voltage that includes noise associated with PWM control is compared with the neutral point voltage, and a signal that is output by the comparator is directly applied to the microcomputer input port. Is assumed to be detected directly. In the assumed prior art, as shown in FIG. 9, a period for polling the input port level is set with a certain width, and within that period (hereinafter, the width of the period is referred to as “detection window”). Polling is performed continuously. The timing at which the output signal level of the comparator is switched from high to low or from low to high is detected as a zero cross point. Such detection of the zero cross point is performed for the non-energized phase.

  When the motor is driven and controlled by energization at 120 degrees, a zero cross point is generated every three electrical phases for three phases. Therefore, as shown in FIG. 9A, when the motor is rotating at a constant speed, the energization pattern is switched when a period corresponding to an electrical angle of 30 degrees elapses from the previous zero-cross point detection timing, and from that timing. If the time point corresponding to the electrical angle of 60 degrees has elapsed within the next detection window, the zero cross point can be detected there. That is, the start point of the detection window is set before the period corresponding to the electrical angle of 60 degrees, and the zero cross point is detected in the vicinity of the start point. The end point of the detection window is until the time when the zero cross point cannot be detected. Keep expanding.

  On the other hand, as shown in FIG. 9B, when the motor is accelerating, the timing at which the zero-cross point arrives next is advanced, so that it cannot be detected before the start point of the detection window. In this case, if it is determined as “detection error”, the starting point of the detection window is set as a zero cross point, and the energization pattern is switched after 30 degrees of electrical angle, and the next detection window is set when 60 degrees of electrical angle has elapsed. Note that the timing of determining “detection error” may be determined when the motor is in an acceleration state and cannot be detected by the midpoint of the detection window. As a result, the zero cross point detected next time is biased toward the start point side of the detection window, so that the followability to acceleration is deteriorated.

  As a result of setting the detection window as described above, when the motor decelerates as shown in FIG. 10, the zero cross point is detected on the start point side of the detection window, so that the followability to the deceleration is good. If the zero cross point cannot be detected in the detection window during deceleration, it is determined that the sensorless drive control has failed and the motor startup sequence is executed again.

  In consideration of the above consideration, in the present embodiment, as shown in FIG. 11A, the detection window is set narrower, and the level of the output signal of the comparator 16 is detected at the start point and end point of the detection window. If one level exceeds the neutral point voltage and the other level is lower than the neutral point voltage, it can be estimated that the zero cross point is within the detection window. In that case, it is estimated that there is a zero cross point at the midpoint of the detection window, and switching to the next energization pattern is performed 60 degrees after the previous energization pattern switching timing. In this case, the detection window is set by setting a start point and an end point before and after the point when 60 degrees has elapsed from the previous zero cross point estimation timing.

  Then, as shown in FIG. 11 (b), when it is determined that there is no zero cross point in the detection window when the motor 14 is accelerating (in this case, it is before the start point), the start point of the detection window is set to the zero cross point. As a reference, the next energization pattern switching is performed after (30-α) degrees have elapsed. Further, the next detection window is set by setting the start point before α degrees and the end point after α degrees so that the zero cross point can be captured in the detection window. Thereby, the followability at the time of acceleration becomes good.

  Further, in the case of deceleration shown in FIG. 12, if it is determined that there is no zero cross point in the detection window (in this case, after the end point), the end point of the detection window is set as the zero cross point, and (30 + α) degrees have elapsed with reference to that point. The next energization pattern switching is performed later. Further, the next detection window is set so that both the start point and the end point are expanded by α degrees as in acceleration. Also in this case, the followability during deceleration is good.

  In FIG. 2, what is surrounded by a thin broken line with the thick broken line as the center is the position of the detection window in the previous detection cycle. On the other hand, the detection window at the tip of the right-pointing arrow is the current detection window in which the energization pattern switching timing is changed and the detection width is expanded when the zero-cross point cannot be detected last time.

  Next, the control for executing the above processing will be described more specifically. In FIG. 3, the microcomputer 2 is interrupted after a predetermined time has elapsed in order to avoid noise generated in synchronization with switching between energization and non-energization of the FETs 12 and 13 from the on timing of the PWM signal. Then, first, it is determined whether or not the time at that time has crossed the zero-cross detection start time (starting point of the detection window) (S1). If it has crossed (YES), the output signal level of the comparator 16 is read (S2, second). 1 detection timing). In the subsequent step S3, it is determined whether or not the current time has crossed the zero-cross detection end time (end point of the detection window) (S3). If it has crossed (YES), the output signal level of the comparator 16 is read again (step S3). S4, second detection timing). Details of the processes in steps S2 and S4 will be described later.

  Based on the execution result of step S4, if it is the detection time of the zero cross point (S4a: YES), the rotational speed of the motor 14 is measured from two detection time intervals (S5), the target rotational speed is calculated (S6), and both The PWM duty is calculated by performing PI control (S7). Thereby, the PWM duty is adjusted so that the rotation speed of the motor 14 converges to the target rotation speed. Then, the output voltage of the amplifier 20 is A / D converted to calculate the read current limiting rate (S8), and the temperature limiting rate is calculated with reference to the terminal voltage of the thermistor 18 (temperature of the module 11) (S9).

  Further, when the microcomputer 2 reads the drive power supply voltage divided by the resistance elements 21 and 21 by A / D conversion, the microcomputer 2 calculates a voltage correction rate (S10). Then, the PWM duty obtained in S7 is corrected by each correction factor calculated in steps S8 to S10, and the rotation speed of the motor 14 when the next PWM interruption occurs is predicted (S11).

  Next, the same determination as in step S4a is performed (S11a), and if it is a zero cross point detection time (YES), a time at which the next zero cross point is predicted to be calculated is calculated (S12). Then, the start point and end point timing (time) of the detection window are calculated based on the time (S13), and the next energization switching timing is calculated from the previous zero cross point (S14). Then, an interrupt to be generated in order to switch the energization pattern next time is set (S15). In step S11a, if it is not the zero cross point detection time (NO), an interrupt generation timer for measuring the PWM period is set (S16).

  As shown in FIG. 4A, in step S2, when the microcomputer 2 reads and acquires the output voltage of the comparator 16 (S21), the microcomputer 2 determines whether or not the voltage is in the detected (1) phase (S22). Here, “detection (1) phase” and “detection (2) phase” to be described later are levels of any one of U, V, and W determined according to each energization pattern as shown in FIG. Corresponds to the levels before and after transition when transitioning to H and L. If the phase is the detection (1) phase (YES), the process is terminated. If the phase is not the detection (1) phase (NO), the current time, that is, the execution time of step S2 is set as the zero cross point detection time (S23). . Then, when the energization angle to be advanced is incremented (S24), the detection of the zero cross point is set to “failure” (S25), and a free-run determination described later is performed (S26).

  As shown in FIG. 4B, in step S4, it is first determined whether or not the time point is detecting a zero cross point (S31). Here, “being detected” is set from the time when the energization pattern is switched. If “detecting” (YES), the output voltage of the comparator 16 is read and acquired (S32), and it is determined whether the voltage is in the detection (2) phase (S33). If it is the detection (2) phase (YES), the zero-cross point is straddled between steps S2 and S4 (detection condition is satisfied), so the estimated time of the zero-cross point is the expected time calculated in step S12. (S38). In this case, the detection (estimation) of the zero cross point is set to “success” (S39), and the process proceeds to step S37 (same free run determination as S26). In the present embodiment, “detection” for the zero-cross point means “estimation”.

  In step S33, if the output voltage of the comparator 16 is not in the detection (2) phase (NO), the current time (execution time of step S4) is set as the zero cross point detection time (S34). Then, the energization angle to be advanced is decremented (S35), and the detection of the zero cross point is set to “failure” (S36), and the process proceeds to step S37.

  In the free-run determination shown in FIG. 5, it is first determined whether or not the detection of the zero cross point is “failure” (S41). If it is “success” (NO), it is determined whether or not the number of detection failures is one or more (S46), and if it has never failed (NO), the process is terminated. If the detection of the zero-cross point is “failure” in step S41 (YES), the counter for counting the number of detection failures is incremented (S42), and whether the number of detection failures has reached a predetermined threshold, for example, 12 times or not. Is determined (S43). In this description, the number of detection failures is used, but instead of the number of detection failures, a time during which the zero cross point cannot be detected may be measured to determine whether or not the time has exceeded a predetermined threshold.

  In step S43, if the number of detection failures is 11 times or less (NO), the free-run determination process is terminated, but if it is 12 times or more (YES), the drive mode of the motor 14 is set to a free-run state (S44). 14 is set to 0 rpm (S45), and the process is terminated. At this time, the energization by the inverter circuit 3 is turned off for all phases, and the generation of an interrupt for switching the energization pattern is prohibited.

  That is, when there is no zero cross point between steps S2 and S4, the motor 14 is driven by open loop control assuming that the start point or end point of the detection window is the zero cross point. If the zero cross point is not detected between steps S2 and S4 even if the state driving is continued, the rotational position of the motor 14 and the energization pattern do not match, or the assumed rotational speed of the motor 14 is It is assumed that it is far from the actual rotational speed. Therefore, it is expected that the PWM control is temporarily stopped and the motor 14 is in a free-running (idling) state, and a zero cross point is detected in a state where the motor 14 is rotating with inertia.

  The case where “YES” is determined in step S46 is a case where the detection has succeeded after having failed once or more in the past. In this case, it is determined whether or not the actual rotational speed of the motor 14 exceeds 0 rpm (S47). If it is not 0 rpm (YES), the process is terminated. On the other hand, if the motor 14 is stopped and the actual rotational speed is 0 rpm (NO), the detection failure frequency counter is cleared to zero (S48), the advance value is set to the initial value (S49), and the motor The drive mode 14 is forcibly commutated and restarted (S50).

Here, in step S13 shown in FIG. 3, the time width of the detection window may be adjusted according to the zero cross point detection result (success / failure) determined in the process shown in FIG. If the detection is successful, the next detection window is narrowed to improve the estimation accuracy. If the detection is unsuccessful, the next detection window is expanded to increase the probability that the next detection will be successful. That is, the control contents shown in FIGS. Further, when the detection window is expanded, the energization time is shortened accordingly, and the time width is secured.
At this time, if the zero-cross point detection failure continues, the energization time is gradually shortened. However, if the lower limit value is set for the energization time and the detection condition is not satisfied even if the energization time reaches the lower limit value, the motor 14 rotates. The number may be reduced or the PWM control may be stopped.

As shown in FIG. 6A, in the period in which the motor 14 accelerates, the detection interval of the zero cross point for every electrical angle of 60 degrees gradually decreases. Then, as shown in FIG. 6B, the energization pattern is switched before the advance angle equivalent time from the timing of the electrical angle of 30 degrees from the zero cross point. The energization pattern counter is incremented at each switching timing, and the count value circulates repeatedly from “0” to “11”.
The energization pattern is 12 patterns by adding six transition periods including the timing at which the six energization patterns shown in FIG. That is, when the count value indicates “1, 3, 5, 7, 9, 11”, “in detection” is set in step S31. As for the advance angle α, when the energization switching timing when the advance angle is not advanced is defined as “energization angle”,
α = (conduction angle−60 degrees) / 2
It will be equivalent time.

  As shown in FIG. 7, the microcomputer 2 refers to the output signal of the comparator 16 corresponding to each of the U, V, and W phases, and detects (estimates) the zero cross point. Then, the actual rotation speed of the motor 14 is calculated by averaging the times when the zero cross point is detected. In addition, the detection result of the zero cross point has three patterns, that is, a case where the detection is successful and a case where the detection point is the start point (early) or the end point (late) of the detection window. Used for averaging and calculation of actual rotation speed. The above detection result is applied to the determination of the sign and the like for the advance angle correction at the time of switching the energization pattern.

  The actual rotational speed of the motor 14 is also used for calculating the energization angle and calculating the energization pattern switching time. As for the energization pattern switching time, the next switching and the next switching time (next time) are calculated in consideration of the detection time of the zero cross point and the advance correction value. When energization is started from a state where the actual rotational speed is 0 rpm in the free run determination, the energization pattern switching time is calculated by replacing the next time with the current time.

  As shown in FIG. 8, when starting the motor 14 from a stopped state, the rotor is first positioned by DC excitation. Then, when energization is stopped and waiting until the current decreases to some extent, rotation is started by forced commutation. During the forced commutation period, it is assumed that there is a zero cross point after a predetermined time from the switching timing of the energization pattern, that is, the next energization pattern is switched at a timing corresponding to the number of rotations expected in the initial stage of rotation. The energization pattern is switched by setting an interrupt, and an interrupt for the next energization pattern switching is set during the interrupt processing.

  When a predetermined time elapses from the start of forced commutation or when the energization pattern is switched by a predetermined number of revolutions, estimation of the zero cross point is started. Alternatively, the zero cross point may be estimated in parallel, and if the zero cross point can be detected before the next energization pattern switching interrupt occurs, the estimation may be continued thereafter. The estimation may be started when the rotation speed of the motor 14 reaches the target value. Thereafter, the PWM duty is adjusted by PI control or the like so that the rotational speed converges to the target value.

  Furthermore, a final target value (threshold value, for example, 3 A) of the current detected by the amplifier 20 is set. Then, when the rotational speed is increased by forced commutation, the target value of the current is increased stepwise, and the PWM duty is corrected by, for example, PI control so that the detected current converges to each target value. The estimation of the zero cross point may be started when the detected current reaches the final target value. This prevents the energization current at the start-up from becoming excessive.

  As described above, according to the present embodiment, the microcomputer 2 estimates the zero-cross point of the induced voltage based on the change in the terminal voltage of the stator winding 15 that is the non-energized phase of the motor 14 and transfers it to the stator winding 15. Is controlled to output a PWM command value. Specifically, the first detection timing (S2) is set after a predetermined time has elapsed from the start time of the PWM cycle, and the second detection timing (S4) is set after the predetermined time has passed.

  If one of the output voltages of the comparator 16 detected at the first and second detection timings is lower than the neutral point voltage of the motor 14 and the other exceeds the neutral point voltage, The middle of the detection timing is estimated as the zero cross point, or the expected zero cross point is assumed, and the first and second detection timings in the next detection cycle are determined based on the estimated zero cross point. On the other hand, if the detection condition is not satisfied, the time between the first and second detection timings in the next PWM cycle is set to be longer.

  That is, when the detection condition is satisfied, the zero cross point is considered to be within the detection window, and therefore, the middle of the time width of the detection window can be reasonably estimated as the zero cross point without directly detecting the zero cross point. On the other hand, if the detection condition is not satisfied, the zero cross point is considered to be outside the window. Therefore, the time width of the window used for the next detection cycle is made longer so that the zero cross point can be captured within the window.

  Therefore, since the microcomputer 2 can set the energization timing based on the zero cross point of the induced voltage without performing the filtering process, the processing load of the CPU control program is reduced as compared with the conventional one, and functions other than the motor control are also performed in parallel. Execution becomes possible. Further, when the estimation of the zero cross point is successful, the time width of the detection window in the next detection cycle is adjusted to be narrowed, so that the zero cross point estimation accuracy can be improved.

  Further, the microcomputer 2 stops the PWM control when the number of times that the zero-cross point estimation failure is continuously satisfied without satisfying the detection condition exceeds a predetermined threshold value, and the motor 14 is put into a free-run state to estimate the zero-cross point. . Therefore, even if the rotational position of the motor 14 does not match the energization pattern, or the assumed rotational speed of the motor 14 is far from the actual rotational speed, the motor 14 rotates due to inertia. It can be expected that a zero-cross point is detected in the state where

  Further, the microcomputer 2 starts the motor 14 by forced commutation, starts estimation of the zero cross point when the current detected via the amplifier 20 exceeds a predetermined threshold, and fixes based on the estimated zero cross point. Since the energization timing to the child winding 15 is controlled, it is possible to smoothly shift from forced commutation to sensorless control. Further, the microcomputer 2 measures the rotational speed of the motor 14 based on the estimated zero cross point, and when the rotational speed reaches the target value, the PWM duty is adjusted so as to converge to the target value as it is, so that the duty is optimized. can do.

Further, when the time between the first and second detection timings is made longer, the microcomputer 2 shortens the energization time to the stator winding 15 and expands the time width of the detection window. If the detection condition is not satisfied even when the energization time reaches the lower limit value, the rotational speed of the motor 14 is reduced or the PWM control is stopped. Therefore, the probability that the zero cross point is detected is further increased, or the motor 14 is turned on. The zero-cross point can be detected again by restarting.
In addition, even when the rotation of the motor 14 is stable, the microcomputer 2 periodically detects the energization current, the drive voltage, and the temperature of the element module 11 and limits the PWM duty based on the detection results. The occurrence of overcurrent and overheating can be prevented.

(Second Embodiment)
As shown in FIG. 13, in the second embodiment, learning control is performed so as to optimize the energization time (energization angle). That is, when the energization angle is calculated as shown in FIG. 7, it is determined whether or not the actual rotational speed of the motor 14 is stable (S51). Specifically, it is determined whether the state where the change in the actual rotational speed is within a predetermined range continues for a predetermined time (for example, about 1 second) or continues for a predetermined number of times of switching the energization pattern. If the motor 14 is accelerating or decelerating, the actual rotational speed is not stable, so it is determined as “NO”, and it is estimated that the zero cross point has changed with the fluctuation of the rotational speed from the already calculated time. (S52).

  On the other hand, if it is determined in step S51 that the actual rotational speed is stable (YES), the degree of change in duty is determined in steps S53, S55, S57, and S59. A predetermined threshold is set for the change in duty, and if the change in which the sign is negative or positive is greater than or equal to the threshold (S53: YES, S57: YES), the energization time is expanded by 1 electrical angle (electrical angle) ( S54, S58).

  If the change on the negative side of the duty is ½ or less of the threshold (S55: YES), the energization time is narrowed by 0.5 degrees (S56), and the change on the positive side of the duty is ½ or less of the threshold. (S59: YES), the energization time is extended by 0.5 degrees (S60). By controlling in this manner, the duty angle is optimized by offsetting the energization angle within a range of ± 1 degree as shown in FIG.

  That is, in the processing in steps S56 and S60, the energization time is intentionally changed for optimization learning. Accordingly, in the case where “YES” is determined in step S53, the rotational speed continues to be stable even if it is varied in step S56. Therefore, the amount of variation is increased in step S54 to accelerate the optimization. The same can be said for the reverse polarity fluctuation in the process of step S58.

  As described above, according to the second embodiment, when the fluctuation of the actual rotational speed of the motor 14 is within a predetermined range, the microcomputer 2 applies the stator winding 15 to the stator winding 15 according to the change degree of the PWM duty in that state. Since the duty is optimized by changing the energization time, the driving efficiency of the motor 14 can be improved. In this case, even if the energization time is changed, if the fluctuation of the rotation speed of the motor 14 remains within a predetermined range, the change amount of the energization time is increased. Therefore, the optimization learning can be accelerated and the optimization learning can be completed in a short time.

The present invention is not limited to the embodiment described above or illustrated in the drawings, and the following modifications or expansions are possible.
The threshold value of the number of detection failures in step S43 is not limited to 12 and may be changed as appropriate.
In 2nd Embodiment, you may change suitably also about the energization angle value increased / decreased by step S54, S56, S58, S60.
The switching element is not limited to a MOSFET but may be a bipolar transistor or IGBT.
The present invention is not limited to controlling a motor mounted on a vehicle, but can be applied as long as it controls a brushless DC motor.

  In the drawing, 2 is a microcomputer (PWM signal output means, control circuit), 3 is an inverter circuit (drive circuit), 14 is a motor (brushless DC motor), 18 is a thermistor (temperature detection means), and 19 is a shunt resistor (current). Detection means), 20 is an amplifier (current detection means), and 21 and 22 are resistance elements (voltage detection means).

Claims (14)

In a motor control device that PWM-controls a brushless DC motor (14) by a position sensorless system,
PWM signal output means (2) for generating and outputting a PWM signal based on the PWM command value;
The CPU is configured to control the energization timing of the stator winding by estimating the zero-cross point of the induced voltage based on the change in the terminal voltage of the stator winding (15) of the brushless DC motor. And a control circuit (2) for outputting the PWM command value,
The control circuit sets a first detection timing for the terminal voltage of the stator winding that is in a non-energized phase after a predetermined time has elapsed from the start of the PWM cycle, and then the terminal voltage after a predetermined time has elapsed. Set the second detection timing of
When one of the terminal voltages detected at the first and second detection timings is lower than the neutral point voltage of the motor and the other of the terminal voltages exceeds the neutral point voltage, the detection condition is satisfied. An intermediate between the first and second detection timings is estimated as the zero cross point, and the first and second detection timings in the next detection cycle are determined based on the estimated zero cross point,
If the detection condition is not satisfied, the time between the first and second detection timings in the next detection cycle is set to be longer , and based on the number of rotations of the motor measured from the estimated zero-cross point interval A motor control device that calculates the PWM command value .   2. The motor control device according to claim 1, wherein when the estimation of the zero-cross point is started, the control circuit adjusts so that the interval between the first and second detection timings in the next detection cycle is narrowed.   The control circuit stops the PWM control when the number of consecutively determined estimation failures without satisfying the detection condition exceeds a predetermined threshold value, and estimates the zero cross point in a state where the brushless DC motor is idling. The motor control device according to claim 1, wherein:   The control circuit stops the PWM control when the time during which the detection condition is not satisfied and the zero-cross point cannot be estimated exceeds a predetermined threshold, and the zero-cross point is estimated in a state where the brushless DC motor is idling. The motor control device according to claim 1, wherein the motor control device is a motor control device. Current detection means (19, 20) for detecting a current supplied to the motor;
The control circuit starts the motor by forced commutation,
When the current detected by the current detection means exceeds a predetermined threshold, the estimation of the zero cross point is started, and the energization timing to the stator winding is controlled based on the estimated zero cross point. Item 5. The motor control device according to any one of Items 1 to 4.   The said control circuit measures the rotation speed of the said motor based on the estimated zero cross point, and adjusts the said PWM command value so that the said rotation speed converges on the said target value. The motor control apparatus as described in any one. If the fluctuation of the rotation speed of the motor is within a predetermined range, the control circuit changes the energization time to the stator winding according to the change degree of the PWM command value in that state , and the fluctuation of the rotation speed 7. The motor control device according to claim 1 , wherein the PWM command value is optimized so as to remain within the predetermined range .   The control circuit further increases the amount of change in the energization time if the variation in the rotation speed of the motor is within a predetermined range even if the energization time to the stator winding is changed. The motor control device according to claim 7.   9. The control circuit according to claim 1, wherein when the time between the first detection timing and the second detection timing is made longer, the energization time to the stator winding is shortened accordingly. The motor control device described in 1.   The motor control device according to claim 9, wherein the control circuit reduces the rotational speed of the motor if the detection condition is not satisfied even when the energization time reaches a lower limit value.   The motor control device according to claim 9, wherein the control circuit stops PWM control of the motor if the detection condition is not satisfied even when the energization time reaches a lower limit value. Current detection means (19, 20) for detecting a current supplied to the motor;
12. The control circuit according to claim 1, wherein the control circuit periodically detects the current by the current detection unit and limits the PWM command value based on the detected current. Motor control device. Voltage detection means (21, 22) for detecting the drive voltage of the motor;
13. The control circuit according to claim 1, wherein the control circuit periodically detects the drive voltage by the voltage detection unit, and limits the PWM command value based on the detected drive voltage. The motor control apparatus described. Temperature detecting means (18) for detecting the temperature of the drive circuit (3) for energizing the stator winding based on the PWM signal;
14. The control circuit according to claim 1, wherein the control circuit periodically detects the temperature by the temperature detection unit and restricts the PWM command value based on the detected temperature. Motor control device.

Images