Method and device for characterising particles

Abstract

The invention relates to a method for characterising particles (6) in an evaluation unit (9) which is configured to receive polarisation-dependent intensity signals (8) which contain information relating to a field distribution of a laser beam (4) through which the particles (6) can be guided. The method comprises the steps of generating a time series of particle signals which are calculated from the received intensity signals (8), reading in or ascertaining a locus curve of defined reference values for intensity and polarisation direction in the field distribution, determining individual particle signals in the time series which correlate with a reference value of the locus curve, wherein the correlated reference values differ in their polarisation directions, and outputting the features of the particles (6). The invention also relates to a sensor arrangement comprising a correspondingly configured evaluation apparatus.

Description

[0001] 11.03.2024 DS17322P3811WO0 Method and device for characterizing particles Description Field of the invention The invention relates to a method for characterizing particles in an evaluation unit comprising a processor unit and a communication interface which is configured to output the features of the particles, wherein the processor unit is configured to receive polarization-dependent intensity signals which contain information of a field distribution of a laser beam through which the particles can be passed,and wherein the field distribution at each position has a defined combination of a local intensity and a local polarization direction of the laser beam. Furthermore, the invention relates to a sensor arrangement with a correspondingly configured evaluation device. Background of the Invention The characterization of particles and particle streams is of high relevance in industrial practice. Parameters of interest in this regard are, in particular, the size and shape of the particles, as well as their position, speed, and direction of movement in a particle stream. Document DE 102019209213 A1 describes a sensor arrangement for characterizing particles in which particles traverse a laser beam in a defined measuring range. The field distribution of the laser beam is designed such that a unique combination of intensity and polarization direction exists for each point in the field distribution.Based on these, relevant particle parameters can be determined, for example, based on Stokes parameters. A possible algorithm for determining the position of a particle in a field distribution with a radially symmetric polarization direction of a laser beam is described in the article by Stefan Berg-Johansen et al.: “Classically entangled optical beams for high-speed kinematic sensing” in Optica, Vol. 2, No. 10 (2015), pages 864-868 (including Supplementary Material). In the proposed algorithm, the Stokes parameters S0, S1, S2, S3, S4, S5, S6, S7, S8, S9, S10, S11, S20, S21, S32, S43, S54, S65, S66, S77, S80, S81, S92, S83, S84, S95, S86, S96, S97, S10, S11, S12, S13, S14, S15, S16, S17, S18, S19, S20, S21, S22, S23, S24, S25, S26, S27, S28, S29, S30, S31, S32, S33, S34, S35, S36, S37, S38, S39, S40, S41, S42, S43, S44, S45, S46, S47, S48, S49, S50, S51, S52, S53, S54, S55, S56, S57, S58, S59, S60, S61, S62, S63, S64, S65, S66, S67, S68, S70, S71, S72, S73, S74, S75, S80, S81, S82, S8S1 and S2 are determined and summarized in a lookup table. Taking into account the experimentally determined statistical noise of the laser beam detectors, the lookup table is converted into a probability distribution. During real-time measurement by the sensor, the current Stokes parameters are determined, based on which the current position of a particle can be determined by comparison with the probability distribution. It has been shown in practice that the known methods still require improvement, particularly with regard to the quality of signal detection, the detection of small particles, and detection at high flow velocities. The object underlying the invention is to provide an improved method and a corresponding device for characterizing particles.which ensure more reliable and faster detection of particles. Solution according to the invention In the absence of any contrary indications, any reference to one (including by indefinite and definite articles), two, or another number of objects is to be understood in such a way that the presence of further such objects is not excluded. The reference numerals in all claims have no limiting effect, but are merely intended to improve their readability. The stated object is achieved by a method for characterizing particles with the features of claim 1 and an evaluation unit with the features of claim 9. The method according to the invention for characterizing particles comprises: receiving polarization-dependent intensity signals in an evaluation unit comprising a processor unit and a communication interface,wherein the polarization-dependent intensity signals contain information of a field distribution of a laser beam through which the particles can be passed, and wherein the field distribution has a defined combination of a local intensity and a local polarization direction of the laser beam at each position, as well as the further method steps: (A) generating a time series of particle signals calculated from the received polarization-dependent intensity signals; (B) reading in or determining a locus of defined reference values ​​for intensity and polarization direction in the field distribution; (C) determining individual particle signals in the time series, each of which has a correlation with a reference value of the locus,wherein the correlated reference values ​​differ in their polarization directions; (D) determining features of the particles based on at least two particle signals of the time series identified in step (C); and (E) outputting the features of the particles via the communication interface. An evaluation unit according to the invention for characterizing particles comprises a processor unit and a communication interface configured to output the features of the particles, wherein the processor unit is configured to receive polarization-dependent intensity signals containing information about a field distribution of a laser beam through which the particles can be passed, and wherein the field distribution has a defined combination of a local intensity and a local polarization direction of the laser beam at each position.wherein the processor unit is further configured to: (A) generate a time series of particle signals calculated from the received polarization-dependent intensity signals; (B) read in or determine a locus of defined reference values ​​for intensity and polarization direction in the field distribution; (C) determine individual particle signals in the time series, each of which has a correlation with a reference value of the locus,wherein the correlated reference values ​​differ in their polarization directions; and (D) determining characteristics of the particles based on at least two particle signals of the time series identified in step (C). The stated object is further achieved by a method for characterizing particles having the features of claim 2 and a sensor arrangement having the features of claim 10. The inventive method for characterizing particles in a sensor arrangement comprising a transmitter with a laser source and a mode conversion device, a measurement volume, a receiver, and an evaluation unit with a processor unit and a communication interface, comprises the steps: i) emitting a laser beam from the laser source into the measurement volume; ii) generating a field distribution of the laser beam in the mode conversion device,wherein the field distribution at each position has a defined combination of a local intensity and a local polarization direction of the laser beam, iii) passing the particles through the measurement volume; iv) at least partially receiving the field distribution of the laser beam emerging from the measurement volume and generating polarization-dependent intensity signals of the received field distribution of the laser beam in the receiver; v) transmitting the polarization-dependent intensity signals from the receiver to the evaluation unit; and the further steps: (A) generating a time series of particle signals calculated from the received polarization-dependent intensity signals; (B) reading in or determining a locus of defined reference values ​​for intensity and polarization direction in the field distribution; (C) determining individual particle signals in the time series, each of which has a correlation with a reference value of the locus,wherein the correlated reference values ​​differ in their polarization directions; (D) determining features of the particles based on at least two particle signals of the time series identified in step (C); and (E) outputting the features of the particles via the communication interface. A sensor arrangement according to the invention for characterizing particles comprises: ^ a transmitter with a laser source for generating a laser beam and a mode conversion device for generating a field distribution of the laser beam, wherein the field distribution has a defined combination of a local intensity and a local polarization direction of the laser beam at each position; ^ a measurement volume,which is irradiated by the laser beam and through which the particles can be passed; ^ a receiver for at least partially receiving the field distribution of the laser beam emerging from the measurement volume and for generating polarization-dependent intensity signals of the received field distribution of the laser beam; and ^ an evaluation unit comprising a processor unit and a communication interface configured to output the characteristics of the particles, wherein the processor unit is configured to receive the polarization-dependent intensity signals of the receiver; wherein the processor unit is further configured to: (A) generate a time series of particle signals,which are calculated from the received polarization-dependent intensity signals; (B) reading in or determining a locus of defined reference values ​​for intensity and polarization direction in the field distribution; (C) determining individual particle signals in the time series, each of which has a correlation with a reference value of the locus, wherein the correlated reference values ​​differ in their polarization directions; and (D) determining features of the particles based on at least two particle signals of the time series identified in step (C). Furthermore, the stated object is achieved by a computer program product comprising instructions which, when the program is executed by the evaluation unit according to the invention, cause the evaluation unit according to the invention to carry out the method according to the invention. The interaction of steps (A) to (D) of the method according to the invention,The evaluation unit and sensor arrangement based on a comparison of time series with reference values ​​of a locus in the field distribution enable particles to be detected more reliably than with the prior art. Especially with noisy data, the false detection rate is lower, and smaller particles can be detected. The time required to determine the particles could also be reduced by the method according to the invention, so that even particle flows with high flow velocities can be reliably characterized. Definitions: The "characterization" of particles is understood to mean the determination of characteristics of the particles or the flow of particles, in particular the size and shape of the particles, the volume of the particles, their position in a two-dimensional or three-dimensional field, their trajectory,Direction of movement and speed of movement. The term "trajectory" is used interchangeably with the term trajectory in this context. According to the usual mathematical definition, a "locus" is understood to be the geometric location of a set of points, for example, a curve or line in two-dimensional space. A locus of values ​​in a spatially two-dimensional field distribution with the orthogonal axes X and Y for the directions of polarization, for example, represents a set of points (Xi, Yi) in the coordinate system of polarization. A "particle" in the sense of the present invention is generally understood to mean a particle, in particular a solid particle or a particle of a liquid. The terms "Stokes parameter" and "degree of polarization" are used in their usual meaning to characterize the polarization state of electromagnetic waves. The Stokes parameters S0, S1,S2 and S3 can be determined, for example, by measuring the radiant power after passing through different polarizers and calculating the parameters from this. In the following, the terms "power" and "intensity" are used interchangeably in this context. The parameter S0 corresponds to the sum of the powers measured after passing through a horizontal and a vertical polarizer. The parameter S1 is the difference between the powers measured after passing through a horizontal and a vertical polarizer. The parameter S2 is the difference between the powers measured after passing through a 45° and a 135° polarizer. The degree of polarization (DOP) for linearly polarized light can be calculated, for example, using the following equation: ^^^ =, ( ^^ ^ + ^ ^ ^)^,^ / ^ ^A "time series" is understood to be a sequence of elements in chronological order. A "time series of particle signals" accordingly means the temporal sequence of particle signals. Preferred embodiments of the invention Advantageous embodiments and further developments, which can be used individually or in combination with one another, are the subject of the dependent claims and the following description. The sensor arrangement according to the invention comprises at least one transmitter, a measuring volume, a receiver, and an evaluation unit. The sensor arrangement can also contain several of the aforementioned individual components. TRANSMITTER The transmitter comprises a laser source and a mode conversion device. In principle, any type of laser is suitable as a laser source. For example, a laser diode with a laser wavelength of 1550 nm (nanometers) and an optical power of 100 mW (milliwatts) can be used as the laser source.Such a laser source can be replaced by a laser with a different laser wavelength without limiting the basic principle of the sensor arrangement. The sensor arrangement and the inventive method for characterizing particles can therefore be implemented with any available laser wavelength and combinations of several laser wavelengths, even outside the visible spectrum. Therefore, the sensor arrangement can be used to characterize particles in a variety of different media (gaseous, liquid, or solid). The transverse mode profile of the laser beam generated by the laser source is usually TEM00, but the laser source can also generate a different mode profile. The laser beam can be focused in the transmitter by means of focusing optics, which have at least one focusing optical element, e.g.a lens, into a focal plane, wherein the field distribution generated by the mode conversion device is imaged into the focal plane. The focal plane is preferably located in the region of the measurement volume through which the particles move. The laser beam can be collimated by the focusing optics and optionally expanded by a telescope in order to adapt the beam cross-section to the mode conversion device downstream in the beam path so that it is optimally illuminated. The focal length of the focusing optical element can preferably be adjusted to a desired, sufficiently large working distance between the transmitter and the particle stream or the particles to be characterized. The focusing optics can have a retrofocus lens. The focusing optics can also have zoom optics for adjusting the focus size in the focal plane.The mode conversion device is configured to generate a field distribution of the laser beam, wherein the field distribution has a defined combination of a local intensity and a local polarization direction of the laser beam at each position. A “defined combination” of a local intensity and a local polarization direction of the laser beam means that the assignment between the local intensity and the local polarization direction is known. A particle moving through the measurement range of the field distribution causes a temporal change in both the intensity and the polarization of the field distribution. Due to the defined assignability of a respective polarization / intensity combination to a sub-range orThe particle position can be determined from a position in the field distribution, and based on the temporal progression of the intensity signals, the particle's temporal progression, i.e., its trajectory, can be determined. Based on the temporal progression of the intensity signals, the particle speed, or its temporal progression, can also be determined. Likewise, the particle size, or its temporal progression, can be determined. Particle size refers to the cross-sectional area of ​​a particle in the focal plane. With this sensor arrangement, particles can be characterized at high speeds in the gigahertz range, meaning that particles can be detected or characterized in real time.In a preferred embodiment, the defined combination of a local intensity and a local polarization direction of the laser beam means that the field distribution of the laser beam is unique at every position, in the sense that the combination of a local intensity and a local polarization direction is different at every position in the field distribution. In a further development of this embodiment, the mode conversion device is designed to generate a field distribution of the laser beam with a radially symmetric polarization direction or to generate a field distribution with a line-shaped, constant polarization direction. With this type of field distribution, each combination of intensity and direction of the polarization vector (corresponding to the polarization direction) only exists at a uniquely defined transverse position of the beam profile.A radially symmetric polarization distribution can be achieved, for example, by the superposition of a TEM01 mode and a TEM10 mode. A line-shaped, constant polarization direction can be achieved, for example, by a field distribution that has radial polarization and in which the intensity increases continuously in one direction. In this way, each point on a line has a different intensity, so that the combination of intensity and polarization direction can be clearly assigned to a point. Points on different lines have a different polarization direction and can therefore also be distinguished from one another. In one embodiment, the mode conversion device is designed as a phase plate, a diffractive optical element, a photonic crystal fiber, or a liquid crystal.In principle, it is possible to impart the transverse phase and locally change the polarization on different optical elements. However, it is generally advantageous to perform both the phase and the polarization change on the same optical element. In a further embodiment, the laser source is designed to generate a pulsed laser beam, preferably with a pulse duration of less than 1 ns (nanosecond), for example as an ultrashort pulse laser, with which laser pulses with pulse durations in the picosecond range can be generated. In principle, the laser source can also be designed to generate continuous wave radiation (cw radiation). It is understood that a laser source can also be used that can be switched between an operating mode with continuous wave radiation and an operating mode with pulsed laser radiation.RECEIVER The receiver is designed to at least partially receive the field distribution of the laser beam emerging from the measurement volume and to generate polarization-dependent intensity signals of the received field distribution of the laser beam. The receiver preferably receives the majority of the field distribution, particularly preferably completely. Optics are preferably provided for receiving the field distribution and generating the intensity signals. Analyzer optics are preferably provided for generating the polarization-dependent intensity signals. To receive the field distribution of the laser, the receiver preferably comprises collimation optics for collimating the laser beam or partial beams of the laser beam before they enter the analyzer optics. The analyzer optics are designed to determine the polarization-dependent intensities.In one embodiment, the analyzer optics has at least one detector for this purpose, which is generally not spatially resolving and can be designed, for example, in the form of a photodiode, e.g., a PIN diode. To determine two or more intensities at different (linear) polarization directions, the analyzer optics can, for example, have a fixed polarizer and a polarization-rotating device, e.g., in the form of a rotating λ / 2 delay plate. In one embodiment, the analyzer optics comprises a beam splitter, in particular a geometric beam splitter, for splitting a beam path of the analyzer optics into a first and second detection beam path. The geometric beam splitter can, for example, be designed in the manner of a beam splitter cube or the like and serve to divide the power of the laser beam in a fixed, predetermined ratio (e.g.,50:50) between the two detection beam paths. In a further development, the first detection beam path comprises a first polarization beam splitter and a first and second optical detector, and the second detection beam path comprises a second polarization beam splitter and a third and fourth optical detector. The optical detectors, e.g. in the form of photodiodes such as PIN diodes, enable sampling rates or resolutions of 10 GHz and more. The use of four detectors has proven advantageous in order to determine the polarization parameters, in particular the Stokes parameters, of the field distribution from the polarization-dependent intensities. The optical detectors are adapted to the wavelength(s) of the laser beam or the partial beams. The optical detectors or the PIN diodes can be designed as free-space diodes or as fiber-coupled diodes (single or multimode).The latter have the advantage of being less influenced by stray light. In a further development, the analyzer optics have a polarization-rotating device for rotating a polarization direction of the laser beam or the respective partial beams of the laser beam by 45° either upstream of the first polarization beam splitter or upstream of the second polarization beam splitter. The polarization-rotating device can, for example, be a suitably oriented λ / 2 retardation plate. To determine the polarization parameters, it has proven advantageous if the four detectors each detect the power or intensity at two mutually perpendicular polarization directions (0° and 90° or 45° and 135°). Rotating the polarization direction by 45° between the two detection beam paths simplifies the determination of the polarization parameters from the polarization-dependent intensities.With the exception of the polarization-rotating device, the two detection beam paths, or reflected and transmitted beam components of the laser beam, are interchangeable in the analyzer optics. In a further development, the optical path length of the analyzer optics from the beam splitter to the first to fourth optical detectors is the same. It has proven advantageous if the optical path length in the analyzer optics or from the beam splitter to the detectors is the same length, so that no time-of-flight differences occur between the polarization-dependent intensity signals determined at the optical detectors. In the case where only one laser beam is used, not only the optical path length in the analyzer optics but also from the focal plane to the four detectors is the same.EVALUATION UNIT According to the invention, the evaluation unit comprises a processor unit and a communication interface for outputting the characteristics of the particles. The evaluation unit can comprise further components or functional blocks, for example storage media such as hard disks or RAMs, interfaces for sending or receiving data or signals, for example network components for wired communication such as LAN or for wireless communication such as WLAN or Bluetooth. PROCESSOR UNIT The processor unit is configured to receive polarization-dependent intensity signals. The intensity signals can be transmitted directly from the receiver or receivers to the processor unit in the evaluation unit. However, the signals can also be transmitted indirectly to the processor unit, for example by being temporarily stored on a data carrier.This procedure is suitable for offline applications in which the evaluation of the signals takes place with a time delay. For online applications and in particular for real-time applications, the signals are preferably transmitted directly from the receiver to the evaluation unit. The intensity signals can be transmitted unchanged from the receiver to the evaluation unit or can be subjected to preprocessing before being transmitted to the evaluation unit. Examples of preprocessing include filtering, averaging, and selecting data sets. The processor unit is configured to carry out steps (A) to (D) of the methods according to the invention. The functions represented by these steps can be provided through the use of shared or dedicated hardware, including, but not limited to, hardware that can execute software.Example embodiments may include microprocessor and / or digital signal processor (DSP) hardware, read-only memory (ROM) for storing software that performs the operations described below, and random access memory (RAM) for storing results. The logical operations of the various embodiments may be implemented, for example, as a sequence of computer-implemented steps, operations, or methods executing in a programmable circuit on a general-purpose computer, or a sequence of computer-implemented steps, operations, or methods executing in a programmable circuit with a specific purpose.Embodiments of the invention may also be practiced in network computing environments having many types of computer system configurations, including personal computers, portable devices, multiprocessor systems, microprocessor-based or programmable consumer electronics, network PCs, minicomputers, or mainframe computers. Embodiments may also be implemented in distributed computing environments where tasks are performed by local and remote processing devices connected through a communications network (either through hardwired connections, wireless connections, or a combination thereof). In a distributed computing environment, program modules may be located in both local and remote storage devices. STEP A Step (A) involves generating a time series of particle signals calculated from the received polarization-dependent intensity signals.The particle signals are preferably calculated in such a way that both the information about the intensity and the information about the polarization direction are contained in the particle signals. In one embodiment, the polarization-dependent intensity signals comprise a plurality of different intensity signals that reflect different polarization directions. The polarization-dependent intensity signals preferably comprise the intensity signals I(0°), I(90°), I(45°), and I(135°), which were obtained from the laser beam with polarization angles of 0°, 90°, 45°, and 135°. In this embodiment, it is preferred that one or more of the Stokes parameters and / or variables derived therefrom, such as the degree of polarization, are calculated as particle signal(s) from the different intensity signals.To create a time series, the detected particle signals are stored in chronological order and provided with an identifier for the time assigned to the respective particle signal. STEP B Step (B) of the method according to the invention comprises reading in or determining a locus of defined reference values ​​for intensity and polarization direction in the field distribution. In one embodiment, the reference values ​​of the locus correspond to values ​​of equal intensity. Preferably, the reference values ​​of the locus correspond to values ​​of equal intensity for a field distribution of the laser beam with radially symmetric polarization directions. In this case, the locus ideally corresponds to a circle around the origin of the radially symmetric polarization field. In reality, the circular shape can be deformed due to interference. The locus can be a coherent, continuous curve.However, it can also be defined discontinuously or in sections. In steps (C) and (D), the locus of the reference values ​​is used to determine the characteristics of the particles from the intensity signals read in, wherein at least two different reference values ​​are used for the determination. The shape of the locus of the reference values ​​is therefore preferably selected such that a particle moving through the field distribution in the measurement volume crosses the locus at least twice. In embodiments in which the particles move approximately in a straight line through the field distribution in the measurement volume, the locus preferably comprises a closed, connected curve, in particular a polygon, an ellipse or a circle, or alternatively at least two partial loci that run essentially parallel to one another and at an angle to the expected trajectory of the particle, in particular two mutually parallel straight lines.The type of field distribution is typically predetermined for the respective application, e.g., by the selection of the laser and the specific mode conversion device. In one embodiment, a suitable locus is determined and specified in advance depending on the respective field distribution. A digital representation of the predetermined locus can be read into the processor unit and used for the subsequent steps (C) and (D). In a further embodiment, the locus of the reference values ​​is determined based on the specific field distribution of the laser beam, preferably at a point in time or in a time period during which no particles are moving through the measurement volume. In one embodiment of the invention, the reference values ​​of the locus are polarization-dependent extreme values ​​of the intensity of the field distribution.Choosing the extreme values ​​of the field distribution as reference values ​​facilitates the identification of individual particle signals in step (C), since extreme values ​​are easy to detect in a time series. A further advantage of choosing the extreme values ​​of the field distribution as reference values ​​is that their geometric position in the field distribution relative to the center of the laser beam is known or can be easily determined. STEP C Step (C) of the method according to the invention comprises determining individual particle signals in the time series, each of which has a correlation with a reference value of the locus curve, wherein the correlated reference values ​​differ in their polarization directions. The term “correlation” in this context means that the pair of the respective particle signal and the respective reference value can be assigned based on at least one common property.This can be a qualitative or a quantitative property. In a preferred embodiment, this is the qualitative property of an extreme value. In this case, a correlation exists if the particle signal calculated from polarization-dependent intensity signals has an extreme value and the reference value for the same polarization direction also has an extreme value. In such a case, agreement between the absolute values ​​of the two extremes is not required to establish a correlation. The individual particle signals in the time series can be determined in different ways. Methods for comparing data sets and data series are known in the art. The key differences between the known methods are assumptions that may have to be made in order to perform the comparison, as well as the computing time required for the comparison.In steps (C) and (D) it is assumed that at least two identified particle signals in the time series correlate with different reference values ​​of the locus curve. A comparison of the particle signals in the time series with the reference values ​​can, for example, be based on calculating a parameterizable shape curve based on knowledge of the field distribution and the locus at reference values. The parameters of this shape curve are then adjusted so that the time series of the particle signals is approximated as closely as possible, e.g., using a least squares fitting method. Since this method requires a relatively high computing power, it is more suitable for applications where computing time plays a minor role. It has been shown that a modified pattern recognition method is more suitable for applications where computing time is a critical factor, e.g., in real-time applications.In one embodiment of the invention, in which the reference values ​​of the locus curve are polarization-dependent extreme values ​​of the intensity, step (C) comprises the determination of extreme values ​​of the particle signals in the time series, taking into account a deadband with a predetermined bandwidth that moves along the time scale and is adaptive with respect to the values ​​of the particle signals. An advantage of this embodiment is that the interplay of extreme value search and the adaptive deadband, which moves along with the signal values ​​in the time series, ensures that relevant information can be reliably detected even from noisy intensity signals.In a further embodiment of the invention, in which the reference values ​​of the locus curve are polarization-dependent extreme values ​​of the intensity, step (C) comprises the determination of extreme values ​​of the particle signals of the time series, the determination of a pattern of the time series which comprises at least two determined extreme values ​​and their relationship to one another, and a comparison of the determined pattern with patterns which result from at least two reference values ​​of the locus curve and their relationship to one another.In a preferred embodiment of the invention, in which the reference values ​​of the locus curve are polarization-dependent extreme values ​​of the intensity, step (C) comprises the following substeps: (C1) determining extreme values ​​of the particle signals of the time series, taking into account a deadband of a predetermined bandwidth that moves along the time scale and is adaptive with regard to the values ​​of the particle signals; (C2) determining a pattern of the time series that comprises at least two determined extreme values ​​from step (C1) and their relationship to one another; (C3) comparing the pattern determined in step (C2) with patterns that result from at least two reference values ​​of the locus curve and their relationship to one another. The bandwidth of the deadband can be adapted to the respective application. For very noisy data, a larger deadband bandwidth is recommended, whereas for less noisy data, a smaller deadband bandwidth is sufficient.In one embodiment, the bandwidth of the deadband is determined from a time series of particle signals calculated from the received intensity signals, which were recorded without particles passing through the measurement volume. In this case, the bandwidth of the deadband is preferably defined as the maximum amplitude of the so-called background noise. When defining the bandwidth of the deadband, further information about the signal curves can also be taken into account. For example, before defining the maximum amplitude, the signal data can be checked for plausibility in order to eliminate outliers, which could potentially lead to an excessively large bandwidth. The determination of the extreme values ​​in step (C1) can be carried out in different ways.The consideration of the deadband when determining the extreme values ​​means that smaller fluctuations in the signals that lie within the bandwidth of the deadband are not recognized as extreme values. In this sense, the deadband has a filter function that suppresses noise signals or ignores them when determining the extreme values. This has the advantage that only relevant values ​​are recognized as potential extreme values. In a preferred variant of the method, which has proven to be robust and advantageous in terms of the required computing time, a successive comparison of the values ​​of the particle signals in the time series with the band limits of the deadband is carried out in step (C1). The comparison begins at time k with the particle signal at time k, where k is a time between the first and last time point in the time series. When starting the evaluation, the starting time is preferably selected as the first time point k (k = 1).The deadband is shifted with respect to its boundaries if the value of the particle signal at time k lies outside the band limits. The violated band limit is set to the value of the particle signal at time k. To determine the extreme values, a check is carried out to determine whether the current value of the particle signal at time k is outside the deadband in the opposite direction to the previous shift direction of the deadband. This can be (a) a minimum and (b) a maximum as the extreme value of the value of the particle signal. In case (a), the value corresponding to the current band lower limit is determined as the extreme value if the value of the particle signal at time k lies above the upper limit of the deadband when the deadband was last shifted towards the band lower limit.In case (b), the value corresponding to the current band upper limit is determined as the extreme value if the value of the particle signal at time k is below the lower limit of the deadband when the deadband was last shifted towards the band upper limit. When shifting the deadband, the bandwidth can be adjusted, for example to take dynamics of signal changes in the time series into account. Preferably, the bandwidth is left unchanged so that when the deadband is shifted, both the band lower limit and the band upper limit are adjusted. STEP (C2) In step (C2) a pattern of the time series is determined which comprises at least two determined extreme values ​​from step (C1) and their relationship to one another. Pattern recognition can be carried out in various ways. In one embodiment of the method, the relationship between the at least two extreme values ​​includes the temporal separation of the extreme values.Preferably, it is also checked whether the extreme value in question has a predetermined time interval from the start of the time series and / or the end of the time series. This interval from the start and / or end of the time series is preferably at least as large as the interval between the two extreme values. In a further embodiment of the method, the ratio of the at least two extreme values ​​includes the difference between the extreme values. It is advantageous if the absolute values ​​of the extreme values ​​found lie within a predetermined maximum deviation. The deviation can, for example, be determined by a relative magnitude of the extreme values, which is the quotient of the smaller of the two absolute values ​​to the larger of the two absolute values. Preferably, the relative magnitude is not less than 0.85, so that the smaller absolute value is not less than 85% of the larger absolute value.In a further embodiment of the method, the relationship between the at least two extreme values ​​includes the difference between the extreme values ​​and the values ​​at times between the extreme values. Individual values ​​or values ​​derived from multiple individual values, such as mean values, can be used to calculate the difference or differences. In an advantageous variant, the quotient of the largest absolute value of the value range between the extreme values ​​and the mean of the two extreme values ​​is calculated. This quotient is preferably no greater than 0.85, so that the largest absolute value of the value range between the extreme values ​​is at most 85% of the mean of the two extreme values. The arithmetic mean is preferably used to calculate the mean.STEP (C3) Step (C3) comprises a comparison of the pattern determined in step (C2) with patterns resulting from at least two reference values ​​of the locus and their relationship to one another. Similar to the pattern recognition in step (C2), the pattern recognition in step (C3) can also be carried out in various ways. In one embodiment, the relationship between the at least two reference values ​​of the locus to one another includes the distance between the reference values, the difference between the reference values, and / or the difference between the reference values ​​or their mean values ​​and the corresponding values ​​between the reference values. Preferred criteria and calculation methods arise analogously from the methods mentioned above with regard to step (C2), wherein the reference values ​​are used instead of the extreme values.The preferred criteria for pattern recognition can also be combined, so that a preferred embodiment includes the relationship of the at least two extreme values ​​to one another in step (C2) and the relationship of the at least two reference values ​​to one another in step (C3), the distance between the extreme values ​​or reference values ​​to one another, the difference between the extreme values ​​or reference values, and / or the difference between the extreme values ​​or reference values ​​or mean values ​​of extreme values ​​or mean values ​​of reference values ​​to the corresponding values ​​between the extreme values ​​or reference values. The comparison of the time series pattern determined in step (C2) with the pattern based on the reference values ​​can advantageously be carried out via a parameter comparison of the patterns, for example using the parameters described above, such as distances, differences, quotients.Similar to the comparison of the particle signals of the time series with the reference values ​​using parameterized shape curves described above, the patterns of the time series and the reference values ​​can also be compared, for example, using a least-squares fitting method. In the case of pattern comparison, a significant saving in computing time results, as the comparison is limited to a significantly smaller data set, since the entire field distribution does not have to be examined. The proposed pattern recognition is therefore particularly suitable for real-time applications. STEP D Step (D) involves determining particle characteristics based on at least two particle signals of the time series identified in step (C).After successfully completing the previous steps (A) to (C), at least two particle signals from the time series are known that correlate with at least two different reference values ​​on the locus curve in the field distribution. It is now possible to combine information from the time domain (time series) with information from a local coordinate system (field distribution) and use this to determine the desired particle properties. The features can be determined in different ways. In a preferred embodiment, the features are determined based on known mathematical, in particular geometric, functions and calculation rules that require little computing time. In one embodiment, the calculation is carried out in the following advantageous sequence. First, the associated polarization angles in the field distribution are determined for the identified particle signals.Subsequently, the coordinates of the particle signals in the field distribution are determined from the polarization angles and with knowledge of the position of the locus in the field distribution, which corresponds to the particle positions in the field distribution. From the particle positions, the course of the particle's trajectory through the field distribution can be calculated, for example, by determining the angle of the particle's trajectory to the coordinate axes of the field distribution. From the knowledge of the trajectory, its intersection points with the coordinate axes of the field distribution can be calculated. Finally, from the particle positions and knowledge of the time information associated with the at least two identified particle signals, the velocity of the particle on the trajectory can be determined. The calculation steps can of course also be carried out in a different order or with different calculation rules.In the exemplary embodiment presented below, concrete calculation rules are specified which are also generally applicable to other exemplary embodiments. STEP E The communication interface is configured to receive the determined characteristics of the particles from the processor unit. The communication interface is further configured to output the determined characteristics of the particles. The output can be done in different ways. In one embodiment, the communication interface outputs the characteristics of the particles to a display device that can be visually perceived by a user, for example to a display or a graphical user interface (GUI). This can be, for example, a portable electronic device such as a smartphone, tablet or laptop, or a stationary device. In a preferred embodiment, the visually perceptible display device is integrated into the evaluation unit.In a further embodiment, the communication interface outputs the characteristics of the particles to a data storage device, wherein the data storage device can be located in the immediate spatial proximity of the communication interface or far away from it, the latter, for example, in the case of cloud storage. The communication interface can use protocols commonly used in the prior art to transmit the data. Communication can take place, for example, via a local area network, a token ring network, the Internet, a corporate intranet, wireless signals of the 802.11 series, a fiber optic network, radio or microwave transmission, etc. The underlying technologies are known to those skilled in the art. Brief Description of the Drawings Further advantageous embodiments are described in more detail below with reference to several exemplary embodiments illustrated in the drawings, to which, however, the invention is not limited.The drawings are to be understood as schematic representations. They do not represent a limitation of the invention, for example with regard to concrete dimensions or design variants, unless the description of the drawings indicates otherwise. They show: Figure 1 a schematic representation of a first embodiment of a sensor arrangement according to the invention; Figure 2 a schematic representation of a field distribution with radially symmetric polarization; Figure 3 a time profile of the degree of polarization (DOP) due to a particle movement through the field distribution according to Figure 2. Detailed description of embodiments of the invention In the following description of preferred embodiments of the present invention, the same reference symbols designate the same or comparable components. If there are several identical components, usually only one is provided with a reference symbol.Figure 1 schematically shows a first embodiment of a sensor arrangement according to the invention for characterizing particles. The sensor arrangement comprises a transmitter 1 with a laser source 2 for generating a laser beam 4 and a mode conversion device 3 for generating a field distribution of the laser beam 4. The laser beam 4 radiates through a measuring volume 5 through which particles 6 can pass. A receiver 7 receives the field distribution of the laser beam 4 emerging from the measuring volume 5. In the receiver 7, polarization-dependent intensity signals 8 of the received field distribution of the laser beam 4 are generated and transmitted to an evaluation unit 9, which comprises a processor unit 10 and a communication interface 11. Generic sensor arrangements containing a transmitter, a measuring volume, and a receiver are known from the prior art.In the document DE 10 2019209213 A1, for example, several different configurations of such sensor arrangements are described in detail. In the example shown below, the laser source 2 and the mode conversion device 3 were configured to generate a field distribution of the laser beam 4 that has radially symmetric polarization directions. Fig. 2 shows the field distribution in the focal plane in the measurement volume 5, which is intended for the flow through of the particles 6. With a radially symmetric polarized field distribution, each local polarization direction of the field distribution is aligned radially to the center of the beam profile of the laser beam 4. The local intensity of the laser beam in the field distribution forms a radially polarized mode. The field distribution shown in Fig.The field distribution shown in Fig. 2 represents a superposition of a TEM01 mode and a TEM10 mode and can be generated by the mode conversion device from a TEM00 mode profile that the laser beam 4 has upon exiting the laser source 2. The intensity is encoded by the brightness. Bright areas correspond to high intensity, dark areas to low intensity. The radially symmetric field distribution according to Fig. 2 therefore has a low intensity in the center of the laser beam, which increases radially outwards, reaches its maximum in the area marked by the dashed circle, and then decreases again further radially outwards. Such a distribution is also pictorially referred to as a "doughnut." The field distribution shown in Fig. 2 has a defined combination of the local intensity and the local polarization direction at each position X, Y of the field distribution and thus of the measuring range.This enables each position X, Y of the field distribution to be clearly assigned to exactly one combination of intensity and polarization direction. Furthermore, Fig. 2 shows the trajectory of a particle 6 as a dashed straight line, along which the particle 6 moves through the field distribution in the measuring volume 5. The reference symbol “v” denotes the velocity vector, which points diagonally from top right to bottom left. The trajectory crosses the regions of high intensity of the field distribution and intersects the circle of highest intensity at the particle positions P1 (x1, y1) and P2 (x2, y2). The angles resulting between the vertical axis through the center and the straight line from the center to the two particle positions are denoted by θ1 and θ2. The angle between the particle trajectory and the vertical axis is denoted by θ. TFurthermore, Fig. 2 shows the point of intersection of the trajectory with the horizontal axis, which is designated as point (x0, 0). In the receiver 7 (not shown in detail), the laser beam 4, after passing through the measurement volume 5, was guided via several beam splitters and polarizers into several analyzer optics. The polarizers were configured so that the laser beams with polarization angles of 0° (horizontal polarization), 90° (vertical polarization), 45° and 135° were received in the analyzer optics. Thus, four different polarization-dependent intensity signals I(0°), I(90°), I(45°) and I(135°) were obtained, which were transmitted from the receiver 7 to the processor unit 10 of the evaluation unit 9. In the processor unit 10, the Stokes parameters S0 = I(0°) + I(90°), S1 = I(0°) – I(90°) and S2 = I(45°) – I(135°) as well as the degree of polarization (DOP) were calculated from the polarization-dependent intensity signals as ^^^ = (^^ ^ + ^^ ^ ) ^,^ / ^ ^calculated. The parameters S1 and S2 were normalized by division with the parameter S0. For the subsequent steps, the degree of polarization was selected as the particle signal. A time series of the degree of polarization was generated and stored in the processor unit 10. Based on the knowledge of the radially symmetric polarized field distribution, the extreme values ​​of the intensity in the field distribution were selected as reference values ​​for the locus curve, which are shown as dashed circles in Fig. 2. Due to the radial symmetry of the polarization, the reference values ​​all have the same intensity value. The locus curve of the reference values ​​can, for example, be stored in a memory of the processing unit 10 and be retrievable for processing in the subsequent method steps.To determine individual particle signals in the time series that correlate with reference values ​​of the locus curve, the extreme values ​​of the DOP time series were first determined as particle signals, taking into account a deadband with a predetermined bandwidth that moves along the time scale and is adaptive with regard to the values ​​of the particle signals. The bandwidth of the deadband was defined as the maximum amplitude of the background noise, which could be determined from a time series of particle signals recorded without particles passing through the measurement volume. The extreme values ​​of the particle signals were determined by successively comparing the values ​​of the particle signals in the time series with the band limits of the deadband, starting with the particle signal at the start time of the time series.The deadband was shifted if the value of the particle signal at time k was outside the band limits by setting the violated band limit to the value of the particle signal at time k and adjusting the non-violated band limit accordingly, so that the bandwidth remained unchanged. As soon as the value of the particle signal at time k was above the upper limit of the deadband when the deadband was last shifted towards the lower band limit, the value of the current lower band limit was identified as the minimum. As soon as the value of the particle signal at time k was below the lower limit of the deadband when the deadband was last shifted towards the upper band limit, the value of the current upper band limit was identified as the maximum.The identified extremes were checked for plausibility by comparing their temporal separation, the difference between their absolute values, and the difference between the mean of the two extremes and the time series values ​​between the extremes against threshold values. A "double-peak signal" was determined from the particle signals as a pattern, as shown in Fig. 3 (real noisy signal in Fig. 3). In the next step, a shape curve for an ideal, noise-free double-peak signal was generated in the processor unit 10. To generate the shape curve for this noise-free double-peak signal, either an ideal light mode can be assumed according to the theory: ^^^. ^^^^^ = ^ ∗ ( ^^ ^ + ^ ^ ^) + ^ ^ where ^ determines the amplitude / signal height and ^ ^ an offset value is taken into account. The two orthogonal electric fields are given by: In these coordinates, ^' is the position parallel along the flight trajectory, ^' is the (orthogonal) distance to the center of the donut beam, and ^ ^^‘ or ^ ^^‘ describe the Gaussian waists (width) of the donut beam in the two corresponding directions. Knowing the previously determined extrema values ​​and calibration curves, the shape curve of the ideal double-peak DOP signal (smooth curve in Fig. 3) can be generated. The height of the maxima with the offset provides information about the amplitude A and the offset ^ ^ . The distance between the maxima determines the scaling of the ^' values, where ^ ^ ' is determined by the position of the minimum. The ratio of the height of the maxima to the minimum provides information about the distance ^' from the center via a calibration curve. ^ ^,^^ and ^ ^,^^are related to the donut diameter. Alternatively, the shape curve of the noise-free double-peak signal can be determined via a calibration measurement of the sensor's real mode. The "quality" is then compared to the actual measured curve ^^^ ^^^^ and the shape curve of the noise-free double-peak signal ^^^ ^^^^^ One possible way to determine the deviation between the two curves is to sum the weighted differences with weighting factor ^: where the sum runs over all points in the curves and the denominator serves for normalization. Finally, the calculated score value is compared with a threshold value. As a result of the comparison, the signal is classified as a "correct" particle signal or rejected as a "false" signal. With knowledge of the times of the two extremes and the spatial distribution of the intensity in the field distribution, in particular the radius r of the locus curve (dashed circle in Fig. 2), the desired characteristics of particle 6 could then be easily calculated. From the parameters S1 and S2 at the times of the extremes, the angles θ1 and θ2 to the particle positions P1 and P2 could be determined according to the following equation: 1 ^ ( ^ ) = − arctan 2 Since the arctan function exhibits discontinuities at angles +45° and -45°, it is calculated for all four quadrants in the position coordinates and corrected accordingly if necessary. This also applies to all subsequent equations that use the arctan function. Knowing the angles and the radius r, the coordinates of the particle positions can be determined as: ^ ^ = − sin(^ ^ ) ⋅ ^ ^ ^ = − cos ( ^ ^ ) ⋅ ^ ^ ^ = − sin ( ^ ^ ) ⋅ ^ ^ ^ = − cos(^ ^ ) ⋅ ^ The angle of the trajectory θ T could be calculated from the position coordinates using the following equation: ^ ^ = arctan Alternatively, the angle of the trajectory θT can also be calculated from the signal sums of the parameters S1(t) and S2(t) according to the following system of equations: ^ ^,^^^ = ^^^(^ ^ (^)) ^ ^,^^^ = ^^^(^ ^(^)) 1 ^ ^ = − arctan 2 The angle of the trajectory θT can also be calculated using both methods, which has the advantage of allowing a comparison between the two calculated values, allowing an error estimate. The trajectory crossing position x0 could be calculated as follows: ^ ^ ^ = ^ − ^ ^ ⋅ (^ ^ − ^ ^ ) ^ ^ − ^ ^ Finally, the particle velocity could be determined from the position coordinates and the times of the extrema − y ^ ) ^ Using the method according to the invention, particles of different shapes and sizes could be reliably detected at various flow velocities. Even with significant noise, the method delivered accurate values ​​and was able to reduce the false detection rate compared to prior art methods. Compared to known methods, the computing time required to evaluate the measured values ​​could be significantly reduced. The features disclosed in the above description, claims, and drawings can be important for the implementation of the invention in its various embodiments, both individually and in any combination.

[0002] List of reference symbols 1 Transmitter 2 Laser source 3 Mode conversion device 4 Laser beam 5 Measuring volume 6 Particle 7 Receiver 8 Polarization-dependent intensity signals 9 Evaluation unit 10 Processor unit 11 Communication interface P1 Particle position P2 Particle position

Claims

Claims 1. A method for characterizing particles, comprising: receiving polarization-dependent intensity signals in an evaluation unit which comprises a processor unit and a communication interface, wherein the polarization-dependent intensity signals contain information of a field distribution of a laser beam through which the particles can be passed, and wherein the field distribution has a defined combination of a local intensity and a local polarization direction of the laser beam at each position, characterized by the method steps: (A) generating a time series of particle signals which are calculated from the received polarization-dependent intensity signals; (B) reading in or determining a locus curve of defined reference values ​​for intensity and polarization direction in the field distribution;(C) determining individual particle signals in the time series, each of which has a correlation with a reference value of the locus curve, wherein the correlated reference values ​​differ in their polarization directions; (D) determining features of the particles based on at least two particle signals of the time series identified in step (C); and (E) outputting the features of the particles via the communication interface.

2. A method for characterizing particles in a sensor arrangement comprising a transmitter with a laser source and a mode conversion device, a measurement volume, a receiver, and an evaluation unit with a processor unit and a communication interface, the method comprising the steps: i) emitting a laser beam from the laser source into the measurement volume;ii) generating a field distribution of the laser beam in the mode conversion device, wherein the field distribution has a defined combination of a local intensity and a local polarization direction of the laser beam at each position; iii) passing the particles through the measurement volume; iv) at least partially receiving the field distribution of the laser beam emerging from the measurement volume and generating polarization-dependent intensity signals of the received field distribution of the laser beam in the receiver; v) Transmission of the polarization-dependent intensity signals from the receiver to the evaluation unit; characterized by the further steps: (A) generating a time series of particle signals calculated from the received polarization-dependent intensity signals; (B) reading in or determining a locus of defined reference values ​​for intensity and polarization direction in the field distribution; (C) determining individual particle signals in the time series, each of which has a correlation with a reference value of the locus, wherein the correlated reference values ​​differ in their polarization directions; (D) determining characteristics of the particles based on at least two particle signals of the time series identified in step (C); and (E) outputting the characteristics of the particles via the communication interface. 3.Method according to claim 1 or 2, characterized in that the reference values ​​of the locus curve are polarization-dependent extreme values ​​of the intensity of the field distribution.

4. Method according to claim 3, characterized in that step (C) comprises the following substeps: (C1) determining extreme values ​​of the particle signals of the time series, taking into account a deadband of a predetermined bandwidth that moves along the time scale and is adaptive with respect to the values ​​of the particle signals; (C2) determining a pattern of the time series that comprises at least two determined extreme values ​​from step (C1) and their relationship to one another; (C3) comparing the pattern determined in step (C2) with patterns that result from at least two reference values ​​of the locus curve and their relationship to one another. 5.Method according to claim 4, characterized in that in step (C1) a successive comparison of the values ​​of the particle signals of the time series with the band limits of the dead band is carried out, starting with the particle signal at time k, where k is a time between the first and the last time of the time series, and where the dead band is shifted if the value of the particle signal at time k lies outside the band limits, so that the violated band limit corresponds to the intensity value at time k, and determining the value as an extreme value. which (a) corresponds to the current lower band limit if the value of the particle signal at time k is above the upper limit of the dead band when the dead band was last shifted in the direction of the lower band limit, or which (b) corresponds to the current upper band limit if the value of the particle signal at time k is below the lower limit of the dead band when the dead band was last shifted in the direction of the upper band limit.

6. Method according to claim 4 or 5, characterized in that the bandwidth of the dead band is determined from a time series of particle signals calculated from the received intensity signals, which were recorded without particles passing through the measuring volume.

7. Method according to one of claims 4 to 6, characterized in that in step (C2) the ratio of the at least two extreme values ​​to one another and in step (C3) the ratio of the at least two reference values ​​to one another determines the distance between the extreme values ​​orReference values ​​to each other, the difference between the extreme values ​​or reference values, and / or the difference between the extreme values ​​or reference values ​​or mean values ​​of extreme values ​​or mean values ​​of reference values ​​to the corresponding values ​​between the extreme values ​​or reference values.

8. Method according to one of the preceding claims, characterized in that the field distribution of the laser beam has radially symmetric polarization directions, and the reference values ​​of the locus curve correspond to values ​​of equal intensity. 9.Evaluation unit for the characterization of particles comprising a processor unit and a communication interface which is configured to output the characteristics of the particles, wherein the processor unit is configured to receive polarization-dependent intensity signals which contain information of a field distribution of a laser beam through which the particles can be passed, and wherein the field distribution has a defined combination of a local intensity and a local polarization direction of the laser beam at each position, characterized in that the processor unit is further configured to: (A) generate a time series of particle signals which are calculated from the received polarization-dependent intensity signals;. (B) reading in or determining a locus of defined reference values ​​for intensity and polarization direction in the field distribution; (C) determining individual particle signals in the time series, each of which has a correlation with a reference value of the locus, wherein the correlated reference values ​​differ in their polarization directions; and (D) determining features of the particles based on at least two particle signals of the time series identified in step (C).

10. A sensor arrangement for characterizing particles, comprising: a transmitter (1) with a laser source for generating a laser beam and a mode conversion device for generating a field distribution of the laser beam, wherein the field distribution has a defined combination of a local intensity and a local polarization direction of the laser beam at each position;a measurement volume through which the laser beam passes and through which the particles can be passed; a receiver for at least partially receiving the field distribution of the laser beam emerging from the measurement volume and for generating polarization-dependent intensity signals of the received field distribution of the laser beam; and an evaluation unit comprising a processor unit and a communication interface configured to output the characteristics of the particles, wherein the processor unit is configured to receive the polarization-dependent intensity signals from the receiver; characterized in that the processor unit is further configured to: (A) generate a time series of particle signals calculated from the received polarization-dependent intensity signals; (B) read in or determine a locus of defined reference values ​​for intensity and polarization direction in the field distribution;(C) determining individual particle signals in the time series, each of which exhibits a correlation with a reference value of the locus curve, wherein the correlated reference values ​​differ in their polarization directions; and (D) determining features of the particles based on at least two particle signals of the time series identified in step (C); 11. A computer program product comprising instructions which, when the program is executed by the evaluation unit according to claim 9, cause the evaluation unit to carry out the method according to claims 1 to 8.

Images