Method for offset correction of a sensor signal of an inertial sensor, in particular acceleration and / or yaw rate sensor for a motor vehicle

Abstract

Method for offset correction of a sensor signal of an inertial sensor, in particular acceleration and / or yaw rate sensor for a motor vehicle, wherein a) on the one hand, a driving situation-dependent offset determination is carried out by recognizing certain predetermined driving situations and determining the offset in these from the sensor signal, b) on the other hand, the temperature is determined at the inertial sensor or in the vehicle and a dynamic temperature-dependent offset component is determined, c) the sensor signal is corrected on the basis of the offsets determined, and the standstill of the vehicle is used and recognized as the driving situation, characterized in that in at least certain standstill situations new values ​​for the temperature and the sensor offset are determined and stored in a memory as a data set consisting of temperature and sensor-specific offset until a predetermined amount of data sets is available, and that from the time the predetermined amount of data sets is reached a newly determined data set overwrites the data set in the memory which is closest in temperature to the newly determined data set or which has the smallest temperature difference between the two directly adjacent data sets.

Description

[0001] The invention relates to a method for offset correction of a sensor signal of an inertial sensor, in particular an acceleration and / or yaw rate sensor for a motor vehicle according to the preamble of claim 1.

[0002] Inertial sensors, especially gyroscopes and acceleration sensors, output a value at rest that is generally referred to as offset error. The offset error is determined by the nature of a sensor and consists of several components: an initial offset, an offset due to aging, and an offset due to temperature.

[0003] The aim of this invention is to reduce the offset error of a sensor through compensation. This object is achieved by the features of the independent claims. Advantageous developments of the invention emerge from the subclaims, whereby combinations and developments of individual features are also conceivable.

[0004] Several options for offset reduction already exist.

[0005] Signal filters (high-pass filters) have the disadvantage that the sensor signal, which is caused by a driving situation, is incorrectly corrected by the filter as an offset error inherent in the sensor.

[0006] Model-based filtering methods, such as Kalman filters, can distinguish the sensor signal caused by a driving situation from the sensor's own offset error. However, these methods are relatively computationally intensive.

[0007] With a driving-situation-dependent offset determination, driving situations are determined in which inertial forces are not present. The sensor's own offset error is determined as the mean value of the sensor output.

[0008] With environment-dependent offset determination, the sensor's own offset error is corrected by correcting the environmental influence through a temperature measurement and taking into account the temperature characteristic of the sensor's own offset error.

[0009] DE 199 10 868 A1 describes a method for compensating for a zero-position drift of a yaw rate sensor of a vehicle stability control system. In a learning mode, the law according to which the zero point of the yaw rate detected by the yaw rate sensor shifts depending on the operating or ambient temperature is determined. For this purpose, the instantaneous yaw rate is determined for different operating or ambient temperatures of the yaw rate sensor, which are determined in particular based on an oscillation frequency of the yaw rate sensor or a temperature sensor, if it can be deduced from an output signal from another sensor, such as a wheel speed sensor or steering angle sensor, that the vehicle's yaw rate is zero.The yaw rate determined in this way, which deviates from zero due to the influence of temperature, is entered into a table and saved as a zero-point offset value depending on the respective operating or ambient temperature of the yaw rate sensor. To compensate for aging-related zero-point offsets, a comparison is also made between an already determined zero-point offset value and a current zero-point offset value for the same or almost the same temperature value, and the already determined zero-point offset value is replaced by the current zero-point offset value. In this case, if a current zero-point offset value has been determined for the same temperature value for which a zero-point offset value already exists, this existing zero-point offset value is replaced by the current zero-point offset value for this existing (i.e., unchanged) temperature value.During an operating mode, the vehicle's yaw rate detected by the yaw rate sensor is then corrected with a zero point offset value that is read from the table depending on the respective operating or ambient temperature of the yaw rate sensor.

[0010] US 2007 / 0005215 A1 discloses a method for correcting a temperature-related offset value of an acceleration sensor of a notebook. To determine a corresponding offset value, it is first determined whether an absolute value of a temperature difference between a measurement temperature measured by a temperature sensor and a stored reference temperature exceeds a predetermined value. If a corresponding exceedance occurs, a second step is used to determine whether the notebook is in a sleep mode based on the acceleration sensor signal. If a sleep mode has been detected, a next step is used to determine whether the notebook is essentially horizontally aligned, i.e., parallel to a plane imaginary perpendicular to the axis of gravity, based on the acceleration sensor signal and an already existing, stored offset value.If this condition is also met, the corresponding acceleration sensor signal is stored as a (new) offset value along with the measured temperature. When determining the notebook's acceleration, the acceleration sensor signal is then corrected using the stored offset value.

[0011] A key concept of the invention is the combined determination of driving-situation-dependent and environment-dependent offsets. Preferably, a learning method is used to determine the sensor's own temperature characteristic. A prerequisite is the presence of a temperature sensor; modern sensors often have such a sensor, or they can be provided with sufficient accuracy via other sensors in the vehicle.

[0012] For the driving-situation-dependent offset determination, the vehicle's position is determined using at least one acceleration sensor and the influence of gravitational acceleration is taken into account. Alternatively, a wheel sensor or rotation sensor on a vehicle's shaft could also be used.

[0013] The driving situation used and recognized for the driving situation-dependent offset determination is the standstill of the vehicle, since the dynamic disturbances are at their lowest here.

[0014] In at least certain downtime situations, new values ​​for the temperature and sensor offset are determined and stored in memory as a data set consisting of the temperature and the sensor's own offset until a predefined set of data sets is available. In general, it is conceivable that every downtime situation is taken into account; alternatively, specific time intervals between downtime situations to be considered can be used.

[0015] Once the specified number of data records has been reached, a newly acquired data record overwrites the old data record in memory that is closest in temperature to the newly acquired data record. In contrast to a first-in, first-out overwrite concept, this preserves the largest possible dynamic range of data records in memory, making the intended interpolation to the current temperature relatively accurate.

[0016] Alternatively, the data set to be overwritten is determined as follows. The data sets are available with increasing or decreasing temperatures. The data set with the smallest temperature difference between the two directly adjacent data sets is replaced. This achieves an even distribution of the data sets by temperature over time.

[0017] In the preferred embodiment, the method consists of the following modules.

[0018] For driving-situation-dependent offset determination, a driving situation (preferably standstill) is detected, and the offset value of a sensor is determined. The offset value of a yaw rate sensor can be calculated by averaging the sensor output. To calculate the offset value of an acceleration sensor, the position of the vehicle is determined, thus taking into account the influence of gravitational acceleration. The temperature inside the sensor is measured.

[0019] A map with a predefined number of data points is created in the device's memory. Each time the device is idle, new values ​​for the temperature and sensor offset are determined and written to the memory as a data set consisting of the temperature and the sensor's own offset. New values ​​are added as long as the map is not completely full. When the map is full, the data set in the map that is closest in temperature to the already determined data set is overwritten. In a further embodiment, the data set with the smallest temperature difference between the two directly adjacent data sets is overwritten. With this process, the sensor's own temperature characteristic is completely determined over time and continuously updated. The two components: initial offset and offset due to aging are corrected in the process.

[0020] The dynamic component of the sensor offset (temperature influence) is implemented when a driving situation-dependent offset determination is not possible. The dynamic component of the sensor offset (temperature influence) is corrected based on the measured temperature and the determined temperature characteristic curve. The current temperature is determined using the temperature sensor. In the map, two data sets, one below and one above the determined temperature, are used to calculate the current offset values ​​based on linear interpolation. At least two data sets must already have been created through previous driving situation-dependent offset determination; otherwise, no dynamic component of the sensor offset (temperature influence) is corrected. If the measured temperature value lies outside the available data points of the applied characteristic curve, the offset value is determined by linear extrapolation.

[0021] In addition, the calculated offset value can be checked for compliance with the sensor-specific limits (absolute offset, offset gradient) and limited if necessary.

[0022] The application of this method has the advantage that cheaper sensors of lower quality can still be used in applications that are critical with regard to offset errors.

Claims

[1] Method for offset correction of a sensor signal of an inertial sensor, in particular acceleration and / or yaw rate sensor for a motor vehicle, wherein a) on the one hand, a driving situation-dependent offset determination is carried out by recognizing certain predetermined driving situations and determining the offset in these from the sensor signal, b) on the other hand, the temperature is determined at the inertial sensor or in the vehicle and a dynamic temperature-dependent offset component is determined, c) the sensor signal is corrected using the determined offsets, and the stationary state of the vehicle is used and recognized as the driving situation, characterized bythat in at least certain standstill situations, new values ​​for the temperature and the sensor offset are determined and stored in a memory as a data set consisting of temperature and sensor-specific offset until a predetermined amount of data sets is available, and that once the predetermined amount of data sets is reached, a newly determined data set overwrites the data set in the memory which is closest in temperature to the newly determined data set or which has the smallest temperature difference between the two directly adjacent data sets. [2] Method according to claim 1, characterized by that for the driving situation-dependent offset determination, the position of the vehicle is determined by means of at least one acceleration sensor and the influence of gravitational acceleration is taken into account. [3] Method according to one of the preceding claims, characterized bythat the data sets from the memory are used to determine the dynamic temperature-dependent offset component and are adapted to the currently determined temperature by means of interpolation as soon as at least two data sets are available. [4] Control unit for a motor vehicle with at least one inertial sensor and programming for carrying out the method according to one of the preceding claims.