The present invention is directed to a sampling method for an asynchronous sensor and an associated asynchronous sensor.
Sensors for detecting and/or measuring physical variables transmit these variables to a central computing and/or control unit. In analog sensors, digitizing is typically performed synchronously by a sampling rate of the analog-digital converter of a central computing and/or control unit predefined by the operating software. Digital sensors perform a digital transmission of the detected and/or measured physical variables. In asynchronous sampling, the sensor data are updated in a different time range than the retrieval of the data from the system via a digital interface. Due to this asynchronous sampling, a temporal fuzziness (jitter) occurs, which reduces the signal quality.
The sampling method according to an example embodiment of the present invention for an asynchronous sensor and the associated asynchronous sensor may provide the advantage in relation thereto that an output rate of sensor data is elevated and thus the jitter, i.e., the temporal fuzziness, is reduced by a programmable interpolation. The signal quality may be improved by the reduction of the jitter and the performance may thus be increased in regard to reducing the triggering times and improving the accident recognition for passenger protection devices such as airbags, seat belt tensioners, or rollover bars, for example. In particular, complex synchronization of the sensor signals with the system clock of a central analyzing unit may be dispensed with by the triggering method hereof, through which the flexibility in selecting the analyzing concept for the sensor and the technology selection may be increased.
The programmable interpolation may be implemented as linear interpolation. In linear interpolation, for example, a difference of two sampling values is multiplied by the ratio of output rate to sampling rate and the result of the multiplication is added to the old sampling value and stored in an output register. Linear interpolation allows for simple implementation.
To increase the output rate, a required parameter is transmitted via a digital interface from a control unit to the output register during an initialization phase, for example. The required parameter may thus be changed as needed and is available at the beginning of a measurement.
Additionally or alternatively, at least one value for the parameter for increasing the output rate may be stored in a memory in the output register, the value of the parameter being set during the initialization phase by changeover commands which are transmitted from the control unit to the output register via the digital interface, for example. This may allow a simple selection of a predefined value for the required parameter to set the output rate.
Furthermore, it is possible to store at least one parameter for increasing the output rate in the output register as a default setting and load it during an initialization phase. The sampling method starts directly in the “linear interpolation” mode, without transmission of a parameter value or a changeover command. By transmitting a corresponding command, this mode may be shut off or changed as needed, for example, to perform a characterization or an analysis or for applications which do not require linear interpolation.
For safety reasons, the set parameter may be locked against programming during normal operation. Programming or changeover of the parameter between the output rates is thus prevented and the set parameter remains locked until the next initialization phase of the sensor.
An exemplary embodiment of the present invention is illustrated in the drawing and is explained in greater detail in the following description.
In asynchronous sampling, the sensor data are updated in a different time range than from when the data are retrieved from the system via a digital interface. Temporal fuzziness (jitter) arises due to this asynchronous sampling, which may reduce the signal quality and thus the performance in regard to triggering times and crash discrimination as a function of the algorithm used. In conventional digital asynchronous sensors, the updating of the sensor data is set to a fixed value, which typically may not be changed, by an applied clock signal or by an internal clock signal. Removing or reducing the jitter by a complex synchronization of the sensors or by increasing the clock signal frequency is conventional for improving the signal quality. However, limits are set on the improvement by the selected measuring principle and/or by the processing technology and by the hardware construction of the system. In a conventional airbag system, a central acceleration sensor has a sampling rate of 128 μs, for example. A central analyzing unit retrieves the data at a rate of 500 μs, for example. The maximum jitter of the signal data would therefore be 128 μs. A schematic illustration of the jitter, i.e., the temporal fuzziness, in asynchronous sensor signal sampling is shown in
According to example embodiments of the present invention, an output rate of an output register should be increased by programmable interpolation of at least two sampling values. To implement this, in an initialization phase of the asynchronous sensor, for example, during the system start, a real physical sampling rate of 128 μs, for example, is set, i.e., the data are updated every 128 μs in a digital output register of the sensor. In addition, the sensor is set by a corresponding command or a default value in a mode in which the provision of the sensor signals in the digital output register is increased by interpolation of two or more real sampling values, e.g., to 1 μs or 0.5 μs. Depending on the algorithm used, the performance in regard to triggering times and crash discrimination is thus increased.
In an alternative exemplary embodiment, the output register includes a memory, in which at least one value for parameter n for increasing output rate t2 is stored. During the initialization phase, one of the stored values for parameter n is activated by a changeover command transmitted from control unit 6 to output register 5, i.e., a changeover of the output rates may be performed by corresponding commands.
In an additional or alternative exemplary embodiment, at least one parameter n for increasing output rate t2 is stored in the memory of output register 5 as a default setting, parameter n being loaded during the initialization phase. The sampling method thus starts directly in the “linear interpolation” mode, for example. By transmitting a corresponding command, this mode may be changed or shut off as needed, for example, to perform a characterization or an analysis or for applications which do not require linear interpolation.
Sensor 1 shown in
Through the sampling method and the associated asynchronous sensor, control units in airbags, for example, may dispense with synchronization of the sensor signals with the system clock of the central analyzing unit, as well as if “jitter sensitive” algorithms are used. The complex implementation of a synchronization thus becomes superfluous and the flexibility in the selection of the analyzing concept for the sensor and the technology selection is increased.
Number | Date | Country | Kind |
---|---|---|---|
10 2005 002 721 | Jan 2005 | DE | national |
Filing Document | Filing Date | Country | Kind | 371c Date |
---|---|---|---|---|
PCT/EP2005/056572 | 12/7/2005 | WO | 00 | 4/24/2008 |
Publishing Document | Publishing Date | Country | Kind |
---|---|---|---|
WO2006/076988 | 7/27/2006 | WO | A |
Number | Name | Date | Kind |
---|---|---|---|
6259084 | Kochis et al. | Jul 2001 | B1 |
20030110509 | Levinson et al. | Jun 2003 | A1 |
Number | Date | Country |
---|---|---|
197 04 132 | Aug 1997 | DE |
100 54 070 | May 2002 | DE |
10 75441 | Mar 1998 | JP |
2003-32132 | Jan 2003 | JP |
Number | Date | Country | |
---|---|---|---|
20090125278 A1 | May 2009 | US |