METHOD FOR DETERMINING THE READ VALUE OF A SAW SENSOR

Description

Technical area

[0001] The invention relates to a method for determining the readout value of a SAW sensor, wherein a plurality of readout signals are transmitted, the associated response signals are detected and a response signal average is formed, and wherein amplitude extreme value positions of the response signals are determined. State of the art

[0002] To determine the readout value of a SAW sensor, multiple readout signals are typically sent, the response signals associated with the respective readout signals are recorded, and a response signal average is calculated from these. From this response signal average, amplitude extreme value positions are determined as the readout value. However, with multiple SAW sensors with different impulse responses, this averaging leads to the problem that the response signals of all SAW sensors are included in the averaging, making reliable evaluation impossible. This problem is exacerbated by signal interference, which occurs particularly in connection with SAW sensors moving rapidly relative to the readout unit. EP 3 182 084 A1 describes a known method for reading a SAW sensor. Description of the invention

[0003] The invention is therefore based on the object of designing a method of the type described above in such a way that a reliable determination of the readout values ​​of several SAW sensors moved quickly relative to the readout unit is possible.

[0004] The invention achieves the stated object in that for each response signal, a response signal position vector is formed from at least one amplitude extreme value position of the response signal as a scalar, after which the response signals of those response signal position vectors are averaged to form a response signal mean value which are located at a predetermined distance from one another in the vector space spanned by all response signal position vectors, and in that a readout value is determined for each response signal mean value.

[0005] The amplitude extreme value position is understood to be the position value on the abscissa associated with a local maximum value or a local minimum value of the amplitude on the ordinate, fundamentally independent of whether the evaluation is performed in the time or frequency domain. In a preferred embodiment, the evaluation is performed in the time domain, and the amplitude extreme value position corresponds to the transit time of a response signal component. If, according to the invention, a response signal position vector is formed from at least one amplitude extreme value position, a vector space results in which response signal position vectors from identical SAW sensors cluster at certain positions.If only those response signals whose response signal position vectors lie within the range of such a cluster position are selected for averaging, strong interference can be easily filtered out and, in particular, the response signals from different SAW sensors can be separated from each other. Particularly advantageous evaluation conditions arise when those amplitude extreme value positions that are particularly characteristic of the individual SAW sensors are used to generate the response signal position vectors.

[0006] Particularly favorable processing conditions arise if a response signal position vector is formed for each response signal from at least two amplitude extreme value positions of the response signal as scalars.

[0007] For example, several two-dimensional response signal position vectors can be formed from two amplitude extreme value positions of the address portion of the response signals that identify the SAW sensors. This has the advantage that the cluster positions in the spanned two-dimensional vector space can be determined particularly resource-efficiently and quickly using known clustering methods.

[0008] Since each response signal position vector is assigned to a response signal, it fundamentally doesn't matter which response signal position vector is used to assign a response signal to a specific SAW sensor. To disregard random clusters of positions without meaningful information content, only the response signals of those response signal position vectors within a specified distance from each other can be averaged to produce a response signal average. The minimum number can be five, for example, but preferably ten.

[0009] For particularly reliable processing conditions, it is recommended that the amplitude extreme value positions are the abscissa values ​​of local maxima on the ordinate.

[0010] In order to enable a particularly advantageous method, the amplitude extreme value positions of the response signal mean can be determined as a readout value.

[0011] For example, the readout signals can have different frequency components in successive time periods. The response signals are transformed from the frequency domain to the time domain before determining the amplitude extreme value positions of the response signals. The response signals are averaged in the frequency domain to form a response signal mean. The response signal mean is then transformed to the time domain, and the amplitude extreme value positions of the response signal mean are subsequently determined in the time domain. By averaging the response signals in the frequency domain and thus in their recorded, unaltered form, any transformation-related errors can be avoided and the quality of the response signal mean, which is then transformed back to the time domain, can be increased.

[0012] Particularly with SAW sensors moving rapidly relative to an evaluation unit, signal interference in the time domain can occur with temporally staggered readout signals of different frequencies. This is because the respective SAW sensor has already moved so far out of the detection range of the readout unit at the time a readout signal is emitted that the reflected response signal can no longer be recorded or can only be recorded incompletely. To better filter out such erroneous response signals, it is proposed that each response signal be divided into several position ranges and that at least one response signal position vector be formed for each position range from extreme amplitude value positions. In this way, an interference in one position range does not influence potentially undisturbed response signal components in another position range.Although here too the response signals of those response signal position vectors can be averaged to form a response signal mean value which are located at a predetermined distance from one another in the vector space spanned by all response signal position vectors, particularly advantageous readout conditions result if only the response signal components which correspond to the position ranges assigned to the respective response signal position vectors are used to form the response signal mean value. Brief description of the invention

[0013] The drawing shows an example of the subject matter of the invention. Fig. 1 is a block diagram of a flow chart of a method according to the invention, Fig. 2 is a response signal in the time domain, and Fig. 3 is the vector space spanned by several response signal position vectors with cluster positions forming. Ways to implement the invention

[0014] A method according to the invention for determining the readout value of a SAW sensor comprises several, in the Fig. 1 Steps shown in a first embodiment. First, response signals 1 are acquired in the frequency domain (step 2) and transformed back into the time domain using an inverse Fourier transform (step 3). Amplitude extreme value positions are then determined from the response signal 4 in the time domain (step 5) and combined as scalars to form a response signal position vector 6. In the vector space spanned by all response signal position vectors 6, those response signal position vectors 6 that are at a predetermined distance from one another are then determined (step 7). For these response signal position vectors 6 that are at a predetermined distance from one another, a response signal mean value 8 is then formed from the response signals 1 assigned to the response signal position vectors 6 (step 9).These response signal mean values ​​8 are again transformed from the frequency domain to the time domain (step 10) and for each response signal mean value 11 in the time domain the amplitude extreme value positions 12 are detected in step 13, after which the readout value 14 can be determined (step 15) and output (step 16).

[0015] In the Fig. 2 For example, a response signal 4 is shown in the time domain. The response signal comprises four local maxima at amplitude extreme value positions, whereby, for example, a response signal position vector is formed from two amplitude extreme value positions t1, t2.

[0016] Fig. 3shows the vector space spanned by the response signal position vectors of several response signals. It can be seen that two cluster positions 17, 18 arise, corresponding to two SAW sensors. If, according to the method according to the invention, a response signal average is calculated for each of these cluster positions 17, 18 from the response signals assigned to the response signal position vectors located at the cluster position, the response signals of both SAW sensors can be easily separated from one another, and at the same time, highly disturbed response signals, for example, at positions 19, 20, can be filtered out.

[0017] If similar disturbances lead to unwanted clustering positions, only the response signals of those response signal position vectors located within a given distance from each other can be averaged to form a response signal mean value, the number of which exceeds a given minimum number.

[0018] In order to be able to filter out erroneous response signals more effectively, it is proposed that each response signal be divided into several position ranges and that at least one response signal position vector be formed for each position range from amplitude extreme value positions.

Claims

1. Method for determining the readout value of a SAW sensor, wherein a plurality of readout signals are transmitted, the associated response signals (1, 4) are detected and a response signal mean value (8) is formed, and wherein amplitude extreme value positions of the response signals (1, 4) are determined, characterized in that for each response signal (1, 4) a response signal position vector (6) is formed from at least one amplitude extreme value position of the response signal (1, 4) as a scalar, after which the response signals (1, 4) of those response signal position vectors (6) lying at a predetermined distance from one another in the vector space spanned by all the response signal position vectors (6) are averaged to form a response signal mean value (8), and in that a readout value (14) is determined for each response signal mean value (8).

2. Method according to claim 1, characterized in that a response signal position vector (6) is formed for each response signal (1, 4) from at least two amplitude extreme value positions of the response signal (1, 4) as scalars.

3. Method according to claim 1 or 2, characterized in that only the response signals (1, 4) of those response signal position vectors (6) lying within a predetermined distance from one another and exceeding a predetermined minimum number are averaged to form a response signal mean value (8).

4. Method according to one of claims 1 to 3, characterized in that the amplitude extreme value positions are the abscissa values of local maxima on the ordinate.

5. Method according to one of claims 1 to 4, characterized in that the amplitude extreme value positions (12) of the response signal mean value (8) are determined as a readout value (14).

6. Method according to one of claims 1 to 5, characterized in that the readout signals have different frequency components in successive time intervals, in that the response signals (1) are transformed from the frequency domain into the time domain before the amplitude extreme value positions of the response signals (1, 4) are determined, and in that the response signals (1) are averaged in the frequency domain to form a response signal mean value (8), after which the response signal mean value (8) is transformed into the time domain and then the amplitude extreme value positions (12) of the response signal mean value (8) are determined in the time domain.

7. Method according to one of claims 1 to 6, characterized in that each response signal (1, 4) is divided into a plurality of position ranges and at least one response signal position vector is formed for each position range from amplitude extreme value positions.