Sensor and methods using a sensor

Abstract

Inductive or magnetic sensor (1) with at least one sensor element (2) for detecting sensor signals and at least one control and evaluation unit (3) for generating application parameters from the sensor signals, and an output unit (4) for outputting at least one object detection signal, wherein the control and evaluation unit (3) has at least one filter (5) for filtering the application parameters and generating and outputting filter results, where the filter (5) is a filter with an infinite impulse response, the filter (5) is an adaptive filter, and the control and evaluation unit (3) has the filter (5) and is configured to form the object detection signal depending on the filter results of the filter (5), characterized in that the filter (5) has at least four filter parameters, wherein the filter parameters at least one filter value as the starting value of the filter (5), at least one filter value as the final value of the filter (5), at least one incremental value of the filter value and at least one filter counter with a number of values ​​to be determined within a filter stage are, wherein the filter (5) is designed to cyclically capture the application parameters at different times according to the number of values ​​to be determined, and to filter with the filter value as the starting value of the filter (5), and to calculate an updated filter result, wherein the application parameters are cyclically acquired by the filter (5) at different times according to the number of values ​​to be determined, and are filtered with the current filter value, wherein the application parameters are cyclically recorded by the filter (5) at different times according to the number of values ​​to be determined and are filtered with the most recent filter value until a final value of the filter (5) is reached, wherein the control and evaluation unit (3) is designed to evaluate the final filtered filter result in the control and evaluation unit (3) for the calculation of the object detection signal.

Description

[0001] The present invention relates to a sensor according to the preamble of claim 1 and a method with a sensor according to the preamble of claim 5.

[0002] Sensors should be suitable for a wide variety of applications while consistently exhibiting high measurement accuracy. To achieve this, a sensor must adapt to the specific application. For this purpose, the sensor determines various application-specific parameters and uses them to calculate the desired output variable.

[0003] This parameter determination can be considered the sensor's calibration process for the application. It is desirable that this calibration process be as fast as possible while maintaining high accuracy, in order to achieve both rapid commissioning and high measurement accuracy.

[0004] A proven method for achieving high measurement accuracy is to filter the measured parameters. The stronger the filtering effect, the higher the measurement accuracy.

[0005] The stronger the filtering effect, the longer the learning process takes, as each filter introduces a time delay. This effect is amplified if there are dependencies between the parameters being determined. Either a weak filtering effect is set, which allows for a faster learning process but results in lower accuracy, or a strong filtering effect is set, which necessitates a longer learning process but enables high measurement accuracy.

[0006] DE 37 86 165 T2 discloses an active vibration control system for reducing vibrations generated by a primary vibration source with a dominant harmonic frequency that can change rapidly, comprising a processor unit; a first low-pass filter device arranged at an output of the processor device and having a fixed cutoff frequency; a reference device for supplying at least one reference signal to the processor device, representing at least one selected harmonic of the primary vibration sources; wherein, during operation, the processor device generates at least one drive signal using the reference signal and supplies the drive signal(s) to a plurality of secondary vibration sources via the first low-pass filter device; and a second low-pass filter device provided at an input of the processor device and having a fixed cutoff frequency.a sensor device provided at one or more locations and which, during operation, samples a vibration field generated at at least one location by the primary and secondary sources and outputs at least one error signal to the processor device via the second low-pass filter device; with a sampling clock oscillator that supplies a constant sampling clock signal to the processor device, such that the reference signal and the error signal(s) are sampled at a constant rate; wherein the processor device has an adaptive response filter device having first filter coefficients to model the delayed and sustained response of the sensor devices to at least one output of the secondary vibration sources over a wide frequency range;wherein the adaptive response filter device adaptively determines second filter coefficients in response to the fault signal(s) and the drive signal(s) using the first and second filter coefficients to reduce the vibration field measured by the sensor device.

[0007] US Patent 2018 / 0048956 A1 discloses an active attention control system and method for a helmet with a rigid shell that spatially separates the interior of the shell from its surrounding environment. The attention control system comprises receiving at least one playback tone signal, generating playback tones within the shell, generating playback tones from these playback tones, and processing at least one ambient noise signal representative of the sound occurring in the shell's surrounding environment to detect the at least one desired noise signal. The generation of the playback noise is stopped when the at least one desired noise pattern is detected.

[0008] DE 197 49 134 A1 discloses an active vibration damping device and a method for identifying a transfer function in the active vibration damping device. A characteristic signal, representing a quantization of at least one sinusoidal wave, is fed by a control unit of the active vibration damping device to a controlled vibration source synchronously with a predetermined output sampling clock in order to generate a characteristic signal there. The control unit reads a residual vibration signal from a residual vibration detector of the active vibration damping device synchronously with a predetermined input sampling clock. After reading the residual vibration signal as a time series for each frequency, an FFT calculation is performed for each time series to obtain a frequency component of the original sinusoidal wave.An inverse FFT calculation is then performed on the result of the composition of each obtained frequency component in order to derive an impulse response as a transfer function.

[0009] US Patent 201410214214 A1 discloses a method for detecting and responding to disturbances in an HVAC system using a noisy measurement signal and a signal filter. The method includes detecting a deviation in the noisy measurement signal, resetting the filter in response to a detected deviation exceeding a noise threshold, filtering the noisy measurement signal using the signal filter to obtain an estimated state value, and determining that a disturbance has occurred when the estimated state value exceeds a disturbance threshold. In some embodiments, the method further includes performing one or more control actions in response to the detection of a disturbance.

[0010] US Patent 201910353023 A1 discloses a method that includes the insertion of a well bottom assembly into a borehole. The well bottom assembly comprises a rotatable control system and a system for correcting and controlling the borehole position. The borehole correction and control system comprises a first set of sensors whose sensors are positioned near ferromagnetic components of a drill string, and a second set of sensors whose sensors are positioned further away from the ferromagnetic components of the drill string than the sensors of the first set. The borehole correction and control system also includes an RSS controller that is operationally coupled to the first and second sets of sensors and adapted to receive measurement data from them. The controller communicates with the rotatable control system.Faulty data from the first sensor set and reference data from the second sensor set are acquired; the faulty data include measurements from the magnetometer and accelerometer about the transverse axis.

[0011] One object of the invention is to provide a sensor or a method with a sensor for determining application-specific parameters, while simultaneously enabling a fast learning process and high measurement accuracy.

[0012] The problem is solved according to claim 1 with a sensor having at least one sensor element for detecting sensor signals and at least one control and evaluation unit for generating application parameters from the sensor signals, and an output unit for outputting an object detection signal, wherein the control and evaluation unit has at least one filter for filtering the application parameters and generating and outputting filter results, wherein the filter is an infinite impulse response filter, the filter is an adaptive filter, and the control and evaluation unit has the filter and is configured to form the object detection signal depending on the filter results of the filter.

[0013] The problem is further solved according to claim 5 by a method with a sensor having at least one sensor element for detecting sensor signals and at least one control and evaluation unit for generating application parameters from the sensor signals, and an output unit for outputting at least one object detection signal, wherein the control and evaluation unit has at least one filter for filtering the application parameters and generating and outputting filter results, wherein the filter is an infinite impulse response filter, the filter is an adaptive filter, and the control and evaluation unit includes the filter and the object detection signal is formed depending on the filter results of the filter.

[0014] The invention describes a method or a sensor with a control and evaluation unit for determining application-specific parameters or application parameters, which simultaneously enables a fast learning process and high measurement accuracy.

[0015] The invention enables a fast and accurate learning process of application-specific parameters at the same time, especially when dependencies exist between the application parameters.

[0016] The basic idea is that the filtering effect is adapted to a given state and the dependencies between the individual application parameters. This assumes that the application-specific parameters do not change, or only change very slowly. For example, the application parameters might only change over a period of hours, days, or weeks.

[0017] An infinite impulse response filter (IIR filter or IIR system) is a specific type of filter used in signal processing. It is a linear, shift-invariant filter, also known as an LSI (linear shift-invariant) system. Depending on the specific filter parameters chosen, this type of filter, unlike finite impulse response filters, can produce an infinitely long impulse response.

[0018] There are various ways to implement rational IIR filters as a network of addition, multiplication, and delay elements. The actual implementation can take place in microcontrollers, digital signal processors, or digital hardware such as FPGAs or ASICs. In principle, various transfer functions can be implemented in all IIR filter structures.

[0019] An adaptive filter in signal processing is a special type of analog or digital filter that can automatically change its transfer function and frequency during operation. For this purpose, the IIR filter is equipped with a tuning network that can modify the filter coefficients according to specific rules.

[0020] According to the invention, the filter has at least four filter parameters, wherein the filter parameters - at least one filter value as the starting value of the filter, - at least one filter value as the final value of the filter, - at least one incremental value of the filter value and - at least one filter counter with a number of values ​​to be determined within a filter stage, wherein the application parameters are cyclically acquired at different times according to the number of values ​​to be determined, and are filtered with the filter value as the starting value of the filter, and an updated filter result is calculated, wherein the application parameters are cyclically acquired at different times according to the number of values ​​to be determined and are filtered with the current filter value, wherein the application parameters are cyclically acquired at different times according to the number of values ​​to be determined and are filtered with the most recent filter value until a final value of the filter is reached, wherein the final filtered filter result is applied in the control and evaluation unit to calculate the object detection signal.

[0021] If the sensor is not yet configured for the application (e.g., in its factory default state), the lowest filtering level, namely the filter value itself, is selected as the filter's starting value. The control and evaluation unit then determines the application-specific parameters n times (n = number of values ​​to be determined within the filter counter's filter level) and filters them using the lowest filter level, i.e., the filter's starting value. The filter parameter "number of values ​​to be determined within the filter counter's filter level" depends on the complexity of the dependencies. The more complex the dependencies or the more hierarchy levels there are, the higher this filter parameter must be set to achieve a steady state before increasing the filtering level. After the application parameters have been determined n times, the filtering level is increased by an incremental value of the filter value.The larger this filter parameter, the faster high accuracy is achieved. However, this also delays the attainment of a steady state. In this filter stage, the application-specific parameters are determined n times until the filter effect is increased again by an incremental value of the filter value. This process is repeated until the maximum filter stage, i.e., the final filter value, is reached.

[0022] In a further development of the invention, the sensor has a storage unit in which the application parameters are stored and which are read in by the control and evaluation unit when the sensor is switched on.

[0023] To prevent unnecessary relearning for the same application after a sensor restart, the control and evaluation unit persistently saves the filter level and application parameters in the memory unit and starts with these parameters and in this filter level after a restart.

[0024] To simultaneously enable rapid relearning when a previously learned sensor is used in a new application with different application parameters, the control and evaluation unit determines the application parameters directly after restarting without any filtering. If these application parameters deviate from the stored application parameters by more than an adjustable tolerance, all stored application parameters are discarded and the filter level is reset to the initial value. The learning process then begins anew.

[0025] According to the invention, the sensor is an inductive or magnetic sensor.

[0026] In pneumatic cylinders, the piston position is usually determined by an encoder magnet attached to the piston, using an inductive or magnetic sensor. The properties of the magnet, such as orientation, magnetization, remanence, or geometry, can vary.

[0027] An inductive or magnetic sensor is used, for example, for the non-contact magnetic detection of the linear relative movement of a sensor magnet along a measuring section. The sensor has, for example, two identical sensor elements, each of which detects at least one component of the magnetic field of the sensor magnet. The sensor elements are arranged at a distance along the measuring section, and each sensor element generates a monotonic position signal with a value within a range for each position along the measuring section of the sensor magnet. The characteristic curves of the position signals exhibit, for example, point symmetry. The distance between the sensor elements is, for example, shorter than the measuring section. The spatial extent of the sensor elements is also, for example, shorter than the measuring section.

[0028] In a further development of the invention, the application parameters are at least - an offset of sensor signals, - an amplification of offset-corrected sensor signals, - a linearity error, - and / or measurement points of the sensor signals.

[0029] One possible magnetic position sensor is one that calculates the relative position of a sensor magnet to the sensor center using, for example, two sensor elements, each with two field components. Since the magnetic field is highly nonlinear, various application parameters must be determined to ensure accurate position calculation. Furthermore, several corrections must be applied during evaluation within the magnetic field to achieve good position accuracy.

[0030] The following application parameters are determined by the control and evaluation unit. These parameters are determined when the magnet passes over various characteristic points within the measuring range. For some parameters, certain conditions must be met before they can be determined.

[0031] The offsets of the sensor signals are determined as application parameters. Both individual signals from the two sensor elements must be as point-symmetrical and coincident as possible. Any offset must be corrected to restore point symmetry.

[0032] The gain of offset-corrected sensor signals is determined as an application parameter. It can happen that the negative portion of the individual characteristic curve has a different gain than the positive portion. This error must also be compensated to restore point symmetry.

[0033] A linearity error is determined as an application parameter.

[0034] A special algorithm is used to determine the position between the sensor elements, which exhibits a (calculable) linearity error. Calculating and compensating for this error further improves the position determination between the elements.

[0035] Application parameters are defined as measurement points from individual characteristic curves. The offset- and gain-compensated individual characteristic curves are approximated via measurement points and thereby linearized.

[0036] The application parameters have, for example, the following conditions.

[0037] To determine the offsets of the sensor signals, a magnet must travel a minimum distance and pass over the center of the sensor.

[0038] To determine the amplification of offset-corrected sensor signals, the magnet must pass over both sensor elements.

[0039] To determine a linearity error, the magnet must pass over the sensor center and at least one sensor element.

[0040] To determine the measuring points of individual characteristic curves, the measuring points are determined at specific positions calculated by the sensor.

[0041] Until the conditions are met, no application parameters are determined. Therefore, some application parameters are determined immediately, while others are only acquired after several magnet movements. This results in a sequence for determining the application parameters. The measurement points can usually be determined first, as there are hardly any prerequisites. At least two movements are required for the initial determination of the raw signal offset. At least one movement is needed for the amplification of offset-corrected sensor signals. Determining the linearity error requires as many as 20 movements. However, this error has only a very minor influence on the measurement points.

[0042] The application parameters therefore have dependencies on each other: The offset of sensor signals has no dependencies and is based solely on raw data.

[0043] The gain of offset-corrected sensor signals depends on the offset of the sensor signals.

[0044] The linearity error depends on the offset of the sensor signals and on the gain of offset-corrected sensor signals.

[0045] The measurement points of the sensor signals depend on the offset of the sensor signals, the gain of offset-corrected sensor signals, and the linearity error.

[0046] For this hierarchy of dependencies, the filter parameter "Filter counter" is set to five, for example. This should ensure that all application parameters are determined at least once before the filter effect is increased.

[0047] For example, the following values ​​are chosen for the filter parameters: Filter value as starting value of the filter = 0, Filter value as final value of the filter = 0.9 Incremental value of the filter value = 0.45, Filter counter with a number of values ​​to be determined within a filter stage = 5.

[0048] In a further development of the invention, the filter result is calculated according to the following rule: Filter result (n) = Filter result (n−1) * Filter value + Measured value (n) (1−Filter−value) where n = Incremental value of the filter (5).

[0049] The measured value is, for example, the currently recorded value of the application parameter.

[0050] The invention is further explained below with regard to its advantages and features, using exemplary embodiments and the accompanying drawing. The figures in the drawing show: Fig. 1. a sensor on a pneumatic cylinder; Fig. 2 two diagrams for filter evaluation; Fig. 3. Two diagrams on the state of the art.

[0051] In the following figures, identical parts are labelled with identical reference symbols.

[0052] Fig. Figure 1 shows a sensor 1 with at least one sensor element 2 for detecting sensor signals and at least one control and evaluation unit 3 for generating application parameters from the sensor signals, and an output unit 4 for outputting an object detection signal, wherein the control and evaluation unit 3 has at least one filter 5 for filtering the application parameters and generating and outputting filter results, wherein the filter 5 is an infinite impulse response filter, the filter is an adaptive filter, and the control and evaluation unit 3 has the filter 5 and is configured to form the object detection signal depending on the filter results of the filter (5).

[0053] For example, sensor 1 is an inductive or magnetic sensor 1 according to Fig. 1.

[0054] In pneumatic cylinders 9, the piston position is usually determined by means of an inductive or magnetic sensor 1 via a sensor magnet 8 attached to the piston 7. The properties such as orientation, magnetization, remanence, or geometry of the magnet can vary.

[0055] An inductive or magnetic sensor 1 is provided, for example, for the non-contact magnetic detection of linear relative motion of the encoder magnet 8 along a measuring section 10. The sensor 1 has, for example, two identical sensor elements 2, each sensor element 2 detecting at least one component of a magnetic field 11 of the encoder magnet 8. The sensor elements 2 are arranged at a distance along the measuring section 10, and each sensor element generates a monotonic position signal with a value within a range for each position along the measuring section 10 of the encoder magnet 8. The characteristic curves of the position signals exhibit, for example, point symmetry. The distance between the sensor elements 2 is, for example, shorter than the measuring section 10. The spatial extent of the sensor elements 2 is also, for example, shorter than the measuring section 10.

[0056] A possible magnetic sensor 1, or position sensor, is a magnetic position sensor that calculates the relative position of a sensor magnet 8 to the sensor center using, for example, two sensor elements 2, each with two field components. Since the magnetic field is highly nonlinear, various application parameters must be determined to ensure the position can be calculated correctly. Furthermore, various corrections must be made during the evaluation in the magnetic field to achieve good position accuracy.

[0057] For example, filter 5 has at least four filter parameters, where the filter parameters - at least one filter value as the starting value of the filter, - at least one filter value as the final value of the filter, - at least one incremental value of the filter value and - at least one filter counter with a number of values ​​to be determined within a filter stage, wherein the application parameters are cyclically recorded at different times according to the number of values ​​to be determined, and are filtered with the filter value as the starting value of filter 5, and an updated filter result is calculated, wherein the application parameters are cyclically recorded at different times according to the number of values ​​to be determined and are filtered with the current filter value, wherein the application parameters are cyclically recorded at different times according to the number of values ​​to be determined and are filtered with the most recent filter value until a final value of the filter is reached, wherein the final filtered filter result is applied in the control and evaluation unit 3 to calculate the object detection signal.

[0058] If sensor 1 is not yet configured for the application (e.g., in its factory default state), the lowest filtering level, namely the filter value, is selected as the filter's starting value. The control and evaluation unit 3 then determines the application-specific parameters n times (n = number of values ​​to be determined within the filter counter's filter level) and filters them using the lowest filter level, i.e., the starting value of filter 5. The filter parameter "number of values ​​to be determined within the filter counter's filter level" depends on the complexity of the dependencies. The more complex the dependencies or the more hierarchy levels there are, the higher this filter parameter must be set to achieve a steady state before increasing the filtering level. After the application parameters have been determined n times, the filtering level is increased by an incremental value of the filter value.The larger this filter parameter, the faster high accuracy is achieved. However, this also delays the attainment of a steady state. In this filter stage, the application-specific parameters are determined n times until the filter effect is increased again by an incremental value of the filter value. This process is repeated until the maximum filter stage, i.e., the final filter value, is reached.

[0059] For example, sensor 1 has a storage unit 6, in which the application parameters are stored and when sensor 1 is switched on, the application parameters are read by the control and evaluation unit 3.

[0060] To prevent unnecessary relearning for the same application after a Sensome restart, the control and evaluation unit 3 persistently saves the filter level and the application parameters in the storage unit 6 and starts with these parameters and in this filter level after a restart.

[0061] To simultaneously enable rapid relearning when a previously learned sensor 1 is to be used in a new application with different application parameters, the control and evaluation unit 3 determines the application parameters directly after restarting without filtering. If these application parameters deviate from the stored application parameters by more than an adjustable tolerance, all stored application parameters are discarded and the filter level is reset to the initial value. The learning process then begins again.

[0062] For example, the application parameters must be at least - an offset of sensor signals, - an amplification of offset-corrected sensor signals, - a linearity error, - and / or measurement points of the sensor signals.

[0063] The following application parameters are determined by the control and evaluation unit 3. These application parameters are determined when the encoder magnet 8 passes over various characteristic points in the measuring range or measuring section 10. For some application parameters, certain conditions must first be met before they can be determined.

[0064] The offsets of the sensor signals are determined as application parameters. Both individual signals from the two sensor elements must be as point-symmetrical and coincident as possible. Any offset must be corrected to restore point symmetry.

[0065] The gain of offset-corrected sensor signals is determined as an application parameter. It can happen that the negative portion of the individual characteristic curve has a different gain than the positive portion. This error must also be compensated to restore point symmetry.

[0066] A linearity error (LinError) is determined as an application parameter.

[0067] A special algorithm is used to determine the position between the sensor elements 2, which exhibits a (calculable) linearity error. Calculating and compensating for this error further improves the position determination between the elements.

[0068] The application parameters are defined as measurement points from individual characteristic curves. The offset- and gain-compensated individual characteristic curves are approximated via these measurement points and thereby linearized.

[0069] The application parameters have, for example, the following conditions.

[0070] To determine the offsets of the sensor signals, a sensor magnet 8 must travel a minimum distance and pass over the sensor elements 2.

[0071] To determine the amplification of offset-corrected sensor signals, the encoder magnet 8 must pass over both sensor elements 2.

[0072] To determine a linearity error short stroke, the encoder magnet 8 must pass over the sensor center and at least one sensor element 2.

[0073] To determine the measuring points of individual characteristic curves, the measuring points are determined at certain positions calculated by sensor 2.

[0074] Until the conditions are met, no application parameters are determined. Therefore, some application parameters are determined immediately, while others are only acquired after several movements of the encoder magnet. This results in a sequence for determining the application parameters. The measurement points can usually be determined first, as there are hardly any prerequisites. For the offset of the raw signals, at least two movements of the encoder magnet along the entire measuring range 10 are required for the initial determination. For the amplification of offset-corrected sensor signals, at least one movement is required. Determining the linearity error even necessitates 20 movements. However, this error has only a very minor influence on the measurement points.

[0075] The application parameters therefore have dependencies on each other: The offset of sensor signals has no dependencies and is based solely on raw data. The gain of offset-corrected sensor signals depends on the offset of the sensor signals. The linearity error depends on the offset of the sensor signals and on the gain of offset-corrected sensor signals. The measurement points of the sensor signals depend on the offset of the sensor signals, the gain of offset-corrected sensor signals, and the linearity error.

[0076] For this hierarchy of dependencies, the filter parameter "Filter counter" is set to five, for example. This should ensure that all application parameters are determined at least once before the filter effect is increased.

[0077] Fig. Figure 2 shows the application parameters for the adaptive filter (5).

[0078] For example, the following values ​​are chosen for the filter parameters: Filter value as starting value of the filter = 0, Filter value as final value of the filter = 0.9 Incremental value of the filter value = 0.45, Filter counter with a number of values ​​to be determined within a filter stage = 5.

[0079] With the adaptive filter effect, the position is determined very precisely after just two passes. With each increase in the filter effect (after pass five and pass ten), the precision of the position determination increases further. The measurement points are also significantly more accurate and stable.

[0080] For example, the filter result is calculated according to the following rule: Filter result (n) = Filter result (n−1) * Filter value + Measured value (n) (1−Filter−value) where n = Incremental value of the filter (5).

[0081] Fig.Figure 3 shows the settling-in of the application parameters for a constant filter value of, for example, 0.85, according to the state of the art. With a constant filter effect, a slow settling-in of the measurement points can be observed, which means that the position only becomes stable after approximately 10 iterations. Reference symbol: 1 sensor 2 Sensor element 3 Control and evaluation unit 4 output units 5 filters 6 storage units 7 pistons 8 Sensor magnet 9 pneumatic cylinders 10 measuring section 11 Magnetic field

Claims

[1] Inductive or magnetic sensor (1) with at least one sensor element (2) for detecting sensor signals and at least one control and evaluation unit (3) for generating application parameters from the sensor signals, and an output unit (4) for outputting at least one object detection signal, wherein the control and evaluation unit (3) has at least one filter (5) for filtering the application parameters and generating and outputting filter results, where the filter (5) is a filter with an infinite impulse response, the filter (5) is an adaptive filter, and the control and evaluation unit (3) has the filter (5) and is designed to form the object detection signal depending on the filter results of the filter (5), characterized by , that the filter (5) has at least four filter parameters, wherein the filter parameters at least one filter value as the starting value of the filter (5), at least one filter value as the final value of the filter (5), at least one incremental value of the filter value and at least one filter counter with a number of values ​​to be determined within a filter stage are, wherein the filter (5) is designed to cyclically capture the application parameters at different times according to the number of values ​​to be determined, and to filter with the filter value as the starting value of the filter (5), and to calculate an updated filter result, wherein the application parameters are cyclically acquired by the filter (5) at different times according to the number of values ​​to be determined, and are filtered with the current filter value, wherein the application parameters are cyclically recorded by the filter (5) at different times according to the number of values ​​to be determined and are filtered with the most recent filter value until a final value of the filter (5) is reached, wherein the control and evaluation unit (3) is designed to evaluate the final filtered filter result in the control and evaluation unit (3) for the calculation of the object detection signal. [2] Inductive or magnetic sensor (1) according to claim 1, characterized by , that the sensor (1) has a storage unit (6) wherein the application parameters are stored in the storage unit (6) and when the sensor (1) is switched on the application parameters are read by the control and evaluation unit (3). [3] Inductive or magnetic sensor (1) according to claim 2, characterized by that the application parameters are at least - an offset of sensor signals, - an amplification of offset-corrected sensor signals, - a linearity error, - and / or measurement points of the sensor signals. [4] Inductive or magnetic sensor (1) according to at least one of the preceding claims, characterized by that the filter result is calculated according to the following rule Filter result (n) = Filter result (n−1) * Filter value + Measured value (n) (1−Filter−value) where n = Incremental value of the filter (5). [5] Method with an inductive or magnetic sensor (1) with at least one sensor element (2) for detecting sensor signals and at least one control and evaluation unit (3) for generating application parameters from the sensor signals, and an output unit (4) for outputting at least one object detection signal, wherein the control and evaluation unit (3) has at least one filter (5) for filtering the application parameters and generating and outputting filter results, the filter (5) is a filter with an infinite impulse response, the filter (5) is an adaptive filter, and the control and evaluation unit (3) has the filter (5) and the object detection signal is formed depending on the filter results of the filter (5), characterized by , that, the filter (5) has at least four filter parameters, wherein the filter parameters at least one filter value as the starting value of the filter (5), at least one filter value as the final value of the filter (5), at least one incremental value of the filter value and has at least a number of values ​​to be determined within a filter stage, where the application parameters are recorded cyclically at different times according to the number of values ​​to be determined, and filtered with the filter value as the starting value of the filter (5), and an updated filter result is calculated. where the application parameters are recorded cyclically at different times according to the number of values ​​to be determined, and filtered with the current filter value, wherein the application parameters are cyclically recorded at different times according to the number of values ​​to be determined and are filtered with the most recent filter value until a final value of the filter (5) is reached, wherein the final filtered filter result is evaluated in the control and evaluation unit (3) to calculate the object detection signal. [6] Method according to claim 5, characterized by , that the sensor (1) has a storage unit (6) wherein the application parameters are stored in the storage unit (6) and when the sensor (1) is switched on the application parameters are read by the control and evaluation unit (3). [7] Method according to at least one of the preceding claims 5 to 6, characterized by , that the filter result is calculated according to the following rule: Filter result (n) = Filter result (n−1) * Filter value + Measured value (n) * (1− Filter− value) where n = Incremental value of the filter (5).

Images