Method and device for starting control of a sensorless, brushless DC motor

Abstract

Method for controlling the start-up of a sensorless, brushless DC motor by means of at least one control device, wherein at least one start-up parameter can be varied by means of the control device and a BEMF can be observed, wherein a variation process of the at least one start-up parameter is started as a function of at least one variable, wherein the process is continued until at least one predefined termination condition is met, wherein a predefined termination condition depends on the interval between two false starts, wherein a false start occurs if no plausible BEMF can be observed, wherein the start-up parameter is at least the current intensity of a start-up current in the windings of the DC motor or the PWM ratio of the supply to the motor phases, wherein a local optimum is determined by means of an analysis of the intervals between two false starts as a function of the currently set operating point during the variation,where the size for starting the variation procedure depends on the distance between two false starts, whereby it is observed how many attempts could be carried out successfully until a false start occurs again, whereby the variation procedure is carried out when the distance falls below a threshold value.,

Description

[0001] The invention relates to a method and a device for starting control of a sensorless, brushless DC motor.

[0002] Brushless direct current motors (BLDC motors) are well known, although sensorless designs that do not have a rotor position sensor are also known. Such a sensorless BLDC motor is usually operated by determining the zero crossings of the BEMF (back-electromagnetic force) of the de-energized phase. However, the observation or meaningful evaluation of the BEMF curves requires a minimum speed of the DC motor. When the DC motor starts up below this minimum speed, the rotor position is unknown. The motor must therefore be operated purely in a controlled manner, i.e. a rotating field is impressed which is characterized in particular by the temporal progression of the rotational frequency and the current strength. The current strength is set, for example, by a duty cycle of a PWM signal.The choice of the time characteristic for these two variables is critical in many applications because the reliable detection of the BEMF depends on a reliably evenly rotating rotor. However, if the current is set too high during start-up, for example, this can lead to undesirable vibrations in the motor, which no longer allow a precise determination of the BEMF during operation. If the current is set too low, the motor will not start at all. Even if optimal time characteristics have been found as parameters for these two variables, it cannot be assumed that they will remain unchanged. In real systems, the load, the coupling of the load to the motor, and the motor itself will change over time in unpredictable ways. Variations in the properties of the load, coupling, and motor can also require different time characteristics for different examples of a system.An unfavorable choice of these time profiles results in unsuccessful motor start attempts (false starts). A false start occurs when no plausible BEMF can be observed.

[0003] DE 44 38 569 A1 discloses a method for starting an electrically commutated single-strand DC motor, wherein the DC motor is equipped with electromagnetic asymmetry. To energize a winding, an associated switching transistor is switched on and off. To start the DC motor in a specific direction of rotation, a first switching transistor is applied with a first starting pulse, and the voltage induced in the windings is tapped. Depending on the voltage induced in the windings, switching pulses following the starting pulse are generated or not generated. If a switching threshold is exceeded, switching pulses following the starting pulse are generated.If the switching wave is not reached, no switching pulses following the first start pulse are generated, in which case a certain pause is maintained, and after this pause has elapsed, a second switching transistor, which is assigned to the second winding, is subjected to a second start pulse.

[0004] From DE 10 2004 003 153 A1 a control device and a control method for controlling a synchronous motor without using a sensor are known.

[0005] Another method for starting a sensorless synchronous motor is known from DE 10 2006 030 362 A1.

[0006] DE 10 2004 013 241 A1 discloses a method for starting a sensorless, electrically commutable DC motor with a permanent magnet excited rotor and a stator carrying a multiphase stator winding. A switching device is provided for energizing the phase windings, which is controlled by a control device.When the rotor is at a standstill and at the beginning of the starting process in the range below a minimum value of the rotor speed, the control device first determines the position of the rotor and then a controlled initial current is supplied to the phase windings of the stator via the switching device, while after the specified minimum value of the rotor speed has been reached, the control device receives position signals derived directly from the third and / or other odd-numbered harmonics of the phase voltages as rotor position signals for self-commutation of the motor and from these supplies control signals to the switching device for current supply to the phase windings in normal operation.

[0007] DE 197 00 479 A1 discloses a method for starting a two-pulse, electrically commutated DC motor designed for operation in a preferred direction. After switching on, the motor's rotor is forced into oscillation around its rest position by a start routine that alternates current flow in one and the other direction of rotation. During the start routine, the motor is monitored to determine whether it has started. If it is determined that the motor has started, the start routine is exited.

[0008] DE 10 2006 026 560 A1 discloses a starting method for a sensorless brushless motor, in which a controlled start is performed using a preset starting current or PWM ratio. The motor is then controlled based on the EMF zero crossings, with error monitoring and appropriate correction performed during start-up.

[0009] The invention is therefore based on the technical problem of creating a method and a device for starting control of a sensorless, brushless DC motor, by means of which the starting parameterization is improved.

[0010] The solution to the technical problem results from the subject matter having the features of patent claims 1 and 7. Further advantageous embodiments of the invention result from the subclaims.

[0011] For this purpose, the device for controlling the start-up of a sensorless, brushless DC motor comprises a control device, wherein at least one start-up parameter can be varied by means of the control device and a BEMF can be observed. A variation process for the at least one start-up parameter is initiated as a function of at least one variable. The variation process continues until at least one predefined termination condition is met. This allows the start-up parameter(s) to be systematically optimized and adapted to changing conditions, for example, in the case of varying loads.

[0012] The starting parameter is at least the current intensity of a starting current in the windings of the DC motor, which can also be determined, for example, by choosing the PWM ratio.

[0013] A predefined termination condition is the exceeding of a distance between two false starts.

[0014] Furthermore, a local optimum is determined by analyzing the intervals between two false starts as a function of the currently set operating point during the variation. This allows a trend to be determined, for example, whether the optimum is more likely to be achieved at higher or lower values ​​of the varied parameter when only one parameter is varied.

[0015] The interval between two failed attempts is chosen as the parameter for starting the variation procedure. This means that the number of successful attempts is observed before another failed attempt occurs. Only when this interval falls below a threshold value is the variation procedure executed.

[0016] In a further preferred embodiment, the amplitude of the starting parameter(s) is increased, with the increase preferably occurring alternately around the initial position, i.e., for example, starting from an initial position for the current I0, a ΔI is subtracted and added, with ΔI increasing with the number of variation steps. Thus, the magnitude of the amplitude of the change in the starting current increases. This change preferably occurs with a constant step size.

[0017] In a further preferred embodiment, a predefined termination condition is a maximum number of variation steps.

[0018] In a further preferred embodiment, after a predetermined number of variation steps, the system returns to the initial parameter set. The purpose of this measure is to increase the availability of the DC motor, since the initial parameter set is not far from the optimum in systems that do not change abruptly.

[0019] In a further preferred embodiment, a variation direction is blocked if related fault events are detected. For example, no variation toward higher starting currents occurs if the monitoring function reports "overcurrent" during variation.

[0020] In a further preferred embodiment, the sizes of the varied start-up parameters are automatically adjusted in case of non-convergence and the method is repeated, for example, the variation range is widened by increasing the step size.

[0021] A preferred application of the invention is in a DC motor for generating hydraulic pressure for actuators in a motor vehicle. Another preferred application is in a DC radiator / fan motor in a vehicle.

[0022] The invention is explained in more detail below using a preferred embodiment. The figures show: Fig. 1 a schematic representation of the error start-up probability over a parameter and Fig. 2 a schematic representation of the variation for one parameter.

[0023] In the Fig. Figure 1 shows a typical curve of the fault start-up probability p_f for a parameter A, such as the start-up current. With n parameters, an n-dimensional probability trough results. The aim must always be to select the start-up parameters so that they are as close as possible to the bottom of the probability trough. It should be noted that the position of the probability trough can change, for example, due to changing loads.

[0024] In the Fig.Figure 2 shows a schematic representation of the variation of the starting current I over the number of starts V, with the step size Δ being shown. It is assumed that parameter A is the current I and has the value I0 at a starting point of the variation process. If the interval between two false starts for this parameter is smaller than a specified threshold value (for example, every fourth start is a false start), the starting current is increased by the step size Δ and a few starts are carried out at the new starting current value I0 + Δ. If the start-up behavior does not improve, i.e. the interval between two false starts remains below the threshold value, the initial value I0 is reduced by the step size Δ. Start-ups are carried out again using this parameter I0 - Δ. If the interval between two false starts remains below the threshold value even with this parameter I0 - Δ, I0 is increased by a further step size, i.e. the parameter is set to I0 + 2Δ.This process continues until a set parameter reaches a distance between two false starts greater than the threshold value or until a specified number of variation steps have been performed. If necessary, the process is then repeated with a larger step size Δ. However, if a certain distance between two false starts is exceeded at a current value of I0 + x * Δ, this parameter is set as the new operating point. Should it then become necessary to restart the process later, this parameter is used as the starting point for the process.

Claims

[1] Method for controlling the start-up of a sensorless, brushless DC motor by means of at least one control device, wherein at least one start-up parameter can be varied by means of the control device and a BEMF can be observed, wherein a variation process of the at least one start-up parameter is started as a function of at least one variable, wherein the process is continued until at least one predefined termination condition is met, wherein a predefined termination condition depends on the interval between two false starts, wherein a false start occurs if no plausible BEMF can be observed, wherein the start-up parameter is at least the current intensity of a start-up current in the windings of the DC motor or the PWM ratio of the supply to the motor phases, wherein a local optimum is determined by means of an analysis of the intervals between two false starts as a function of the currently set operating point during the variation,where the size for starting the variation procedure depends on the distance between two false starts, whereby it is observed how many attempts could be carried out successfully until a false start occurs again, whereby the variation procedure is carried out when the distance falls below a threshold value., [2] Method according to claim 1, characterized by that the start-up parameter(s) are gradually increased and / or decreased. [3] Method according to one of the preceding claims, characterized by that a predefined termination condition is a maximum number of variation steps. [4] Method according to one of the preceding claims, characterized by that after a predetermined number of variation steps, the system returns to the initial parameter set. [5] Method according to one of the preceding claims, characterized bythat a variation direction is blocked if related error events are detected. [6] Method according to one of the preceding claims, characterized by that if the procedure does not converge, the sizes of the varied start-up parameters are automatically adjusted and the procedure is repeated. [7] Device for controlling the start-up of a sensorless, brushless DC motor, comprising at least one control device, wherein at least one start-up parameter can be varied by means of the control device and a BEMF can be observed, wherein a variation process of the at least one start-up parameter is started as a function of at least one variable, wherein the variation process is continued until at least one predefined termination condition is met, wherein a predefined termination condition depends on the interval between two false starts, wherein a false start occurs if no plausible BEMF is observable, wherein the start-up parameter is at least the current intensity of a start-up current in the windings of the DC motor or the PWM ratio of the supply to the motor phases,wherein a local optimum is determined by analyzing the distances between two false starts as a function of the currently set operating point during the variation, wherein the variable for starting the variation process depends on the distance between two false starts, wherein the device is designed such that it is observed how many starts could be carried out successfully until a false start occurs again, wherein the variation process is carried out when the distance falls below a threshold value. [8] Device according to claim 7, characterized by that the start-up parameter(s) are gradually increased and / or decreased. [9] Device according to one of claims 7 or 8, characterized by that a predefined termination condition is a maximum number of variation steps. [10] Device according to one of claims 7 to 9, characterized bythat after a predetermined number of variation steps, the system returns to the initial parameter set. [11] Device according to one of claims 7 to 10, characterized by that a variation direction is blocked if related error events are detected. [12] Device according to one of claims 7 to 11, characterized by that if the procedure does not converge, the sizes of the varied start-up parameters are automatically adjusted and the procedure is repeated.

Images