Method and device for non-contact temperature measurement of a conveyed material flow, and a granulation device equipped with such a non-contact temperature measurement device

Abstract

The present invention relates to a method and device for non-contact measurement of the temperature of fibrous and / or granular objects in a flow of conveying material, for example at the dryer outlet of a granulator, in which an infrared sensor is used to capture the radiation field of the flow of conveying material flowing against a background, and the temperature of the fibrous or granular objects is determined from a measurement signal from the infrared sensor, wherein the temperature of the background through which the flow of conveying material to be measured passes is varied over time and / or position by a temperature control device, the degree of fluctuation is determined for different background temperatures from the measurement signal from the infrared sensor, the position of minimum fluctuation is determined from these degrees of fluctuation, and the temperature of the object is determined from the value of the measurement signal at the time / position of minimum fluctuation.

Description

[Technical Field]

[0001] The present invention relates to a method and device for non-contact temperature measurement of strangular and / or granular objects in a stream of conveyed material, in which an infrared sensor is used for temperature measurement to measure the temperature of the particles or strangular objects in the stream of conveyed material. The invention also relates to the use of such a non-contact temperature measuring device at the outlet of a pelletizing device, in particular at the outlet of a pellet dryer or strand pelletizer of a pelletizing device. Summary of the Invention

[0002] Accurate temperature measurement is both difficult and important in the processing of granules and their preliminary and intermediate products. If the strands or granules being processed are too hot or too cold, problems will occur in the process, so accurate temperature measurement is necessary, and the temperatures at various points need to be known as accurately as possible in order to accurately control the process. Various control variables such as melt temperature, water temperature, water flow rate, air flow rate, cooling water nozzle flow rate, granule or strand surface area to volume ratio, residence time, etc. can be adjusted in upstream process sections to specifically change the temperature of the object at the measurement point. It can be taken into account that many of these control variables have a linear correlation with the temperature of the object at the operating point, and that changes only take effect at the measuring point after a certain delay. Actuation variables and response times are not an issue for temperature control. In some cases, it is not even necessary to control to a set temperature, but to maintain a temperature range (Temperaturfenster), since other important process variables also need to be controlled. However, in the past, the lack of reliable and accurate temperature measurements meant that temperature control over a temperature range was not possible, but the present invention aims to change this.

[0003] Pelletization processes can be used to produce plastic pellets, typically by forcing a plastic melt through a nozzle-like orifice to produce strands of plastic material. Depending on the pelletization technique, the resulting strands can be cut directly by rotating knives at the outlet of a die plate, as known in underwater pelletization or dry pelletization. In underwater pelletizing, the pellets or granules are dried in a downstream dryer. Alternatively, in strand pelletizing, strands of plastic material first pass through a cooling section in strand form and are then fed into a strand pelletizer where the strands are cut into pellets between a stationary cutting strip and a rotating cutting rotor. However, such pelletizing processes are not only used for plastics or plastic melts, but also in the pharmaceutical sector for the production of tablets or pills, or in the food sector. To achieve a high-quality product, precise temperature control of the processed mass is required.

[0004] For example, an incorrect temperature or cross-sectional temperature profile of the strand that has not yet been pelletized may impair the cutting quality, so accurate temperature measurements of the plastic or material strand at various positions between the die plate and the strand pelletizer may be necessary. By measuring the longitudinal temperature gradient or temperature measurements at two or more evaluation points along the flow of the conveyed material, conclusions can be drawn about the condition of the core of the strand that does not consist of the hot, low-viscosity melt for cutting. On the other hand, if the temperature of the granules deviates significantly from the specified temperature range, it can also cause process problems after grinding. For example, crystallizable plastic pellets or granules require a temperature high enough to trigger an energy-efficient self-crystallization process through their inherent heat, but not too high to prevent the plastic granules from sticking together. In particular, the pellets must be at a predetermined temperature at the outlet of the pellet dryer before being fed into the corresponding downstream processing line or station, such as a vibrating conveyor or a reaction vessel.

[0005] However, measuring the temperature of such strands or granules is difficult for various reasons, and sufficient accuracy has not been achieved to date. On the one hand, this is due to the fundamental problem of measuring the temperature on a moving object. Strands or granules of plastic material move while the temperature is being measured. For example, a slow, dense stream of granules will not flow past the outlet of a centrifugal dryer, while granules swirling around in the air flow will fly past more or less. Depending on the material, the pellets must remain fluid during the temperature measurement process; otherwise, agglomerates may form. This is especially true for plastic pellets, especially those made from sticky plastics, but also for pellets for pharmaceuticals or food products.

[0006] On the other hand, the strands or granules whose temperature is to be measured are often very small, so the sensor system used must be highly sensitive and highly dynamic in order to react sufficiently to the correspondingly small amounts of heat and radiation emitted by the small object, especially considering the high speed of the granules flowing through the sensor system and the rapid lateral vibrations of the strands excited by the granulation process. Depending on the material, the diameter of the strands or granules is often only a few millimeters or a fraction of that, e.g., less than 7 mm, often less than 4 mm, making the object very small compared to the size of the detection range of the infrared sensor.

[0007] In this respect, contact temperature sensors, which are placed in the product stream together with a thermocouple, have been used up to now. However, the problem is that the amount of heat transferred from the product stream or the flowing object to the thermal sensor is usually low. When the object contacts the thermal sensor, the Hertzian contact surface is very small, the contact time is very short, and the thermal conductivity of the material, which is usually plastic, is very low. In contrast, the surface of the thermal sensor continuously radiates heat to the environment while simultaneously absorbing ambient thermal radiation and engaging in convective exchange with the air. Looking at a conventional pellet dryer equipped with a negative pressure fan, the outflowing pellets emerge from the outlet with a boundary layer of heated, moist air. Dry, cool ambient air is drawn against this flow, flowing countercurrently into the dryer through the outlet. The ratio of the cool, countercurrent swirling airflow to the heated air carried with the pellets depends on many parameters and is nearly impossible to adjust reproducibly. Therefore, the contact temperature measurement result is often ultimately just a measurement derived from the air temperature of the various air streams, the product temperature, and the pipe temperature of the outlet pipe.

[0008] Furthermore, non-contact temperature measurements are also being attempted, for example, using infrared sensors for plastic particles in the temperature range of 20 °C to 150 °C. These sensors evaluate the radiative emission of the granules or strands, for example, in the wavelength range of 8 to 14 μm. Pyrometers and bolometric infrared cameras can be used here. The fastest sensors currently available have a response time of just under 10 ms, meaning that a granule or particle passing the measurement spot of an infrared sensor will generate a measurement signal that can be recognized as a peak, but the residence time at the measurement spot is not long enough to allow individual pixels of a pyrometer or bolometer infrared camera to fully control the particle or granule temperature. Similarly, for vibrating strands, the width is so narrow that they are only partially captured by the pyrometer or pixel, and the maximum temperatures of the individual peaks are unacceptably and almost unusably significantly lower than the actual object temperature measured by temporary accumulation or aggregation. As already mentioned, such compression of the product stream is not permitted due to the risk of agglomeration, especially in the case of strands or plastic pellets, and is only possible under laboratory conditions or for test purposes, but not in the operational process of a large-scale plant.

[0009] Direct infrared measurement is only possible if a compact slip packing forms in the exit pipe or chute at high product fill levels, which is not desirable or possible for many products. Measuring the temperature of the strands presents a similar problem: their thinness requires very high resolution infrared cameras, which are very expensive, or multiple cameras placed in the scanner beam, which is very costly.

[0010] The use of powerful infrared sensors such as pyrometers or bolometers with only average resolution results in high signal noise and long response times that prevent the sensor from being perfectly tuned to the pellet temperature. The particle either passes in front of the sensor or passes the measurement spot too quickly for the sensor to respond properly, both of which ultimately stem from Boltzmann's law, which states that the radiant power reaching the sensor system depends on the particle's area and temperature, as well as the measurement distance. More precisely, P Sensor =ε×σ×A×T 4 / r 2 where ε is the emissivity of the plastic, σ is the Stefan-Boltzmann constant, A is the area of ​​the granule, T is the temperature of the granule, and r is the distance between the sensor system and the granule. For example, the temperature of the pellet in the range of 40°C to 120°C is very low for an infrared sensor, and the radiative area of ​​the granule or strand is also very small. Furthermore, the distance between the sensor system and the pellet size is usually very large. Therefore, 1 / r 2 The distance dependence results in a very weak measurement signal. Due to the small size of the granules and strands, low-resolution infrared cameras have the problem that the pixels only partially receive the radiation from the object. Only by continuously and completely covering the pixel can the actual granule or strand temperature be asymptotically approximated and the signal noise reduced at the same time if the granule or strand is present long enough near the sensor system. This is not the case in large-scale processes with high throughput rates.

[0011] Patent document DE10 2016 115 348 A1 attempts to achieve non-contact temperature measurement of glass fiber strands using a thermal imaging sensor, and in order to compensate for the problem of the movement of very thin strands relative to the rather large pixels of the thermal sensor and the small temperature difference of the strand compared to the background, the measurement signal of the thermal sensor is integrated over a longer period and the integral formed is compared with a reference value that can be determined from a corresponding longer measurement of the background, which does not allow the fibers to pass through. A black emitter is used as the background emitter, the reflectivity of which must be at least approximately zero to avoid reflection of thermal radiation in the thermal image of the sensor system. The integral formation is used to determine a so-called average value, which must be more accurate than the maximum value of the thermal sensor's measurement signal at a specific time. However, due to contamination problems, the use of black emitters permanently placed near the strands is hardly manageable in practice in large plants. On the other hand, if process control fluctuations cause the actual temperature of the measurement background to change compared to the reference measurement, more or less large inaccuracies in the required reference measurement will inevitably occur, as is often the case in large-scale processes.

[0012] Patent document WO2014 / 090994 A2 attempts to measure the temperature of a metal strand coated with an insulating plastic sheath without contact, using a radiation sensor that performs spatially resolved thermal radiation measurements from the inside of a tube through which the metal strand passes, the tube also being configured as a black cavity radiator, and with a sufficiently long tube length edge losses can be minimized, allowing measurements under mirror inclusion conditions in the center. This document is based on the finding that when the temperatures of the metal strand and the cavity radiator are identical, the metal strand is no longer visible against the background formed by the inner wall of the pipe, and therefore no significant deviations occur in the area of ​​the moving metal strand in the spatially resolved thermal sensor image. On the one hand, this knowledge is used to infer the temperature deviation of the metal strand compared to the known pipe temperature from the deviation of the radiation sensor signal compared to a reference measurement at a known pipe temperature. On the other hand, this knowledge can be used to control the temperature by adjusting a cavity radiator, configured as a tube and whose temperature is easily measured, to the required target temperature and adjusting process parameters that affect the metal strand temperature when the measurement signal from the thermal sensor exceeds or falls below the corresponding target value, which exists at the metal strand temperature corresponding to the tube temperature. This approach allows the strand temperature to be adjusted to the set temperature, but does not allow the strand temperature to be measured when it is not at the set temperature. Only an uncalibrated measurement of the deviation from the background temperature is determined, i.e., slightly higher or lower than the background. If the cross-sectional area and occupancy density of the object being measured are unknown, quantitative calibration of the object temperature is impossible. Temperatures that do not correspond to the target temperature cannot be accurately measured. Therefore, in particular, it is not possible to monitor whether the temperature range is maintained.

[0013] The underlying object of the present invention is to create an improved method and an improved device of the type described above, which avoids the drawbacks of the prior art and further develops it in an advantageous manner. Preferably, the deviation from the background temperature itself is not only determined but also quantified, so as to be able to output or display the actual object temperature in terms of an absolute temperature value.

[0014] In particular, infrared sensors, with their limited resolution and response time, must enable sufficiently accurate non-contact temperature measurements even in conveying streams containing small, fast-moving objects, such as plastic strands in strand pelletizers or plastic pellets in large-scale pelletizing systems. This is especially true in short system sections, such as the exit of a pellet dryer, where material can move at high speeds without significant packing and remain adequately fluidized. Known solutions for infrared measurements under specular inclusions often cannot be used in large systems due to the long installation lengths, and special solutions are required. The temperature measurement results must be suitable for temperature control, especially for maintaining a temperature range. To keep the control technology simple, it is desirable to obtain a low-noise, quasi-continuous measurement signal with a measurement delay time that is as stable as possible and known.

[0015] According to the invention, this problem is solved by a method according to claim 1, an apparatus according to claim 21 and a granulating apparatus according to claim 34. Preferred embodiments of the invention are set out in the dependent claims.

[0016] Therefore, it is proposed to tackle the signal fluctuations, so to speak, and to examine the intensity of the measurement signal fluctuations in more detail. The source of the fluctuations is the alternation of infrared radiation between the measurement object and the background. Surprisingly, the intensity of the fluctuations in the sensor signal with the simultaneous change in the radiation background can be used to determine what the sensor measures if the granular object is actually present in the sensor's measurement spot long enough, or if the strands are so wide that partial coverage of the sensor pixels is not an issue. In this respect, the background temperature through which the flow of transported material to be measured passes is varied over time and / or position by a temperature control device, and the intensity of the signal fluctuations in the measurement signal from the infrared sensor that occur during the process are evaluated by an evaluation device. In the simplest version, the minimum of the signal fluctuations can be searched for, and the infrared measurement signal at these minimum fluctuations directly represents the temperature of the object. At the minimum of the infrared contrast, the infrared sensor receives the same intensity of infrared radiation from the object as from the background due to its spectral sensitivity characteristics, so the measurement signal directly reflects the temperature of the object.

[0017] This measurement signal is basically suitable for process control and temperature range monitoring in industrial plants, but with this simple method, the background is purely time-varying, so the measurement result is only updated at certain points in time, and measurement noise makes the local fluctuation minima uncertain, which can lead to smaller jumps. This requires the design of simple control techniques that are relatively conservative and slow. More dynamic control requires the training of more complex predictive models. For simple dynamic control of a process within a temperature range, a quasi-continuous measurement signal is desirable.

[0018] Furthermore, accurate temperature measurement by the infrared sensor itself presents challenges: the temperature of the object being measured may differ only slightly from the temperature of the infrared sensor's housing, and the sensor element also experiences thermal radiation from its own electronics and housing. The transmission of conventional infrared optics is also highly temperature-dependent. Therefore, manufacturers of infrared sensors make great efforts to eliminate such influences from the measurement signal using compensation methods with internal temperature monitoring sensors and correction curves. This compensation does not work perfectly, especially in the case of temperature gradients in the measuring head, and deteriorates with age.

[0019] Advantageous embodiments therefore provide means for improving the accuracy of the infrared sensor, such as temperature control of the sensor head, a method for reducing the influence of emissivity without the need for complex two-color pyrometry, an additional temperature measurement of the background by an independent measurement method for online temperature compensation of the measurement signal (online updating of the temperature compensation value is possible even with a material flow or during production interruptions), a calibration station with a blackbody radiator in which the infrared sensor can be placed for short periods of time, etc.

[0020] Furthermore, various statistical methods are included that, on the one hand, robustly quantify signal fluctuations but, on the other hand, also remove the influence of measurement noise and provide low-noise temperature variables for regression models using reference temperatures calculated against the measurement background. The regression model allows for reliable and accurate quasi-continuous determination of object temperature, since the minimum can be found by interpolation or extrapolation with reliable and verifiable quality.

[0021] Furthermore, various devices are described that are suitable for controlling the temperature of the background and the enclosure so that the measurement object is placed in an infrared contrast situation that changes over time and / or position relative to the measurement background, thereby making it possible to measure the object under nearly specular containment conditions in a limited installation space. When analyzing local temperature differences and longitudinal temperature gradients of an object in a measurement area, the locally varying background can also be time-varying and / or multiple evaluation sections can be defined, in which case the temperature of the object and the exact measurement position are determined locally for each section using individual regression models.

[0022] The central starting point for temperature measurement or determination is the analysis of the fluctuations in the infrared sensor signal and the change of these fluctuations relative to the background temperature, with the infrared sensor being positioned in line with the flow of the transported material so that it also receives infrared radiation from the background, at least for a short time or for a portion of the measurement field.

[0023] When evaluating the spectral sensitivity characteristics of an infrared sensor, if there is an infrared contrast between the measurement object and the background, in the case of a fluidized, fast-moving granular object, the object density of the measurement point or group of measurement points varies at each moment, and therefore the infrared measurement signal exhibits highly dynamic fluctuations over time. On the other hand, the flow of material conveyed in strand-like objects occurs in the longitudinal direction, and the strands can move laterally either smoothly or with strong fluctuations depending on their position and mechanical guidance. The infrared contrast between the background and the strands and infrared sensors in the form of line-scan or area-scan cameras produces local intensity fluctuations, especially in the transverse direction of the strands. One pixel may be dominated by the background in the image, while another pixel is dominated by the object, which may be smaller than the resolution of the infrared camera, but the associated sensor pixels are nevertheless controlled in different ways. In the case of strand vibrations, highly dynamic intensity variations with respect to time of the measurement signal also occur, comparable to strand-like or granular objects.

[0024] When evaluating the spectral sensitivity characteristics of an infrared sensor, if there is no infrared contrast between the measurement object and the background, the infrared sensor can stabilize to the temperature of the object for a longer period of time, thereby minimizing temporal and / or positional variations in the measurement signal. Infrared radiation from the background will act on the sensor element with the same intensity as infrared radiation emitted from the measurement object. If the measurement is carried out in an environment that creates specular conditions with uniform infrared radiation from the black emitter from all spatial directions at the object temperature, it does not matter if the emissivity of the object being measured is not ideally ε=1, but for example ε=0.9. In this case, the object emits only 90% of its infrared radiation. However, the missing 10% is fully compensated for by specular reflection from adjacent objects or similarly strongly emitting environments, so that the total radiation is 100%, and therefore the emissivity has no effect on the measurement. Furthermore, the response time of the sensor, the pixel fractional coverage, the surface density of the object, and the dynamics of the movement of the measured object have virtually no effect on the measurement signal at the minimum value of the infrared contrast that stabilizes to the object temperature.

[0025] The two situations with and without infrared contrast differ only in the degree of infrared contrast. Infrared contrast is measured by how different the infrared radiation from the background is from the infrared radiation from the object. If the temperature radiation of the background is higher or lower than the temperature radiation of the object, there will be fluctuations in the measured signal. If the temperature is the same, the radiation is also the same and the fluctuations will be minimal. Since the cause of the fluctuations is intermittent or regional changes in infrared radiation from the object and background, the strength of the fluctuations is also roughly proportional to the temperature difference between the object and background. The larger the temperature difference, the greater the fluctuation, and the smaller the temperature difference, the smaller the fluctuation, until the temperature difference disappears and the fluctuation reaches its minimum value.

[0026] Therefore, to analyze these same measurement signal fluctuations, it is advantageous for the method of measuring the fluctuation intensity to maintain linearity of the fluctuation intensity with respect to the temperature difference. Additionally, variability calculations can be used to reduce large raw data sets of measured signal values ​​to a few meaningful figures. To be able to determine the low-noise fluctuations of the sensor signal relative to the background temperature, it is advantageous to use large amounts of sensor data that can be analyzed using statistical methods. When using a pyrometer as an infrared sensor, which always provides one measurement, as a one-pixel infrared camera does, this data is usually recorded over a certain period of time (time window) and the temporal fluctuations are evaluated.

[0027] Using a line-scan infrared camera aligned with the strands or granules against a uniformly calibrated background, the variability can be calculated directly from each exposure from an evaluation area that includes the entire line-scan, for example. Most data can be obtained with an area infrared camera, especially if the background is calibrated as a temperature gradient field. Here, several different local evaluation zones can be defined, allowing different variability measurements to be determined from a single infrared image for different background temperatures. The evaluation can also summarize several exposures taken under comparable conditions. Since each evaluation zone has its own time and / or positional range, the respective variability can be determined from a larger amount of measurement signal data and with less noise.

[0028] The combination of data in the assessment ranges is based on the following considerations: For example, a certain amount of measurement signal data is required to be able to use statistical methods to determine the degree of variability. All measurement signal data is recorded under similar infrared contrast conditions, especially with a similar background temperature. There are many basic phenomena in the measurement signal that result in more or less radiation from the object, but the details are not relevant for the purpose of temperature measurement and can be described by a few important values ​​in the evaluation range. -The variability of the infrared measurement signal (e.g. signal amplitude, see below for various other options) -Background reference temperature (e.g., background temperature or the reference temperature below) - Position (especially the position in the direction of the flow of the transported material in the spatial center of the evaluation area) -Time (especially the average value between the start and end of the recording) - Key figures from the frequency distribution of the temperature measurement signal, in particular the following figures: Average value of temperature measurement signal Statistical parameters o Other significant figures for the maximum value of the distribution, if applicable.

[0029] The design of the local boundaries of the evaluation bins can be as complex as necessary to cover only the data points in a narrow background temperature range as accurately as possible, but if these characteristic values ​​are subsequently used as input data for a regression model that does not suffer from noisy characteristic values, there is little point in spending a lot of effort to draw the boundaries of the evaluation bins. For highly inhomogeneous adjusted backgrounds, it is easy to define a grid of evaluation regions. In extreme cases, the evaluation region can be considered as a single data point. In this case, the variability is the measured signal value itself. This is because the regression can be calculated even for degenerate forms of this evaluation region. In a direct minimum search without regression modeling, it is only necessary to include multiple data points in the evaluation zone if the background temperature control varies only with respect to time. Apart from these rather theoretical extreme cases, it is particularly useful to use evaluation zones to reduce the measurement signal data to a few characteristic values ​​at an early stage.

[0030] In the following, the intensity of the fluctuations can be understood as a measure that is substantially linearly correlated with the temperature difference between the object and the background. The signal amplitude or span range of the measurement signal is generally suitable as such a measure. To increase robustness against outliers, a small portion of the maximum and minimum measurement data can be omitted, and the decile range or a differently trimmed range can be used as a measure. The interquartile range is usually less suitable because informative data points are excluded as outliers. The standard deviation has proven to be a particularly suitable measure of linear variation because it is less sensitive to individual outliers than signal amplitude, takes into account all data points, and can be easily calculated without internal rank sorting.

[0031] The evaluation zone is characterized in particular by the background temperature at which the measurement signal data is recorded. The boundaries of the evaluation zone are selected so that the difference in background temperature is negligible. Therefore, it makes sense to determine the average background reference temperature of the evaluation zone. In some embodiments, this reference temperature is hereinafter referred to as the reference temperature, and the background temperature may become the reference temperature, since a metrologically determined background temperature is not available.

[0032] To be able to use the reference temperature as a reference variable in the regression model, it is necessary to avoid nonlinear distortions with respect to the background temperature and to be able to use a simple function in the regression model. Therefore, the reference temperature preferably meets the following requirements: a) The reference temperature should be substantially linearly related to the background temperature. In particular, the higher the background temperature, the higher the reference temperature should be, although the reference temperatures may differ by a positive scaling factor. b) In situations where the infrared contrast is at a minimum and the infrared sensor cannot distinguish between the infrared radiation from the object and the background, the reference temperature must assume the actual background temperature, i.e. the object temperature as accurately as technically possible.

[0033] Furthermore, it is advantageous if the values ​​of the reference temperatures of the time and / or position evaluation zones, which are characterized by maintaining a narrow temperature range relative to the background temperature, are determined as noiselessly and stably as possible.

[0034] There are various options for determining the reference temperature of the time and / or location evaluation zone, which differ in particular in the effort required for the sensor system and mathematical modeling. If the evaluation zone covers a period of data acquisition, it is always assumed that the determination of the reference temperature is subjected to time averaging or equivalent filtering, which will not be mentioned below. If necessary, data from neighboring evaluation zones can be used to reduce noise, and advanced regression models can be used to model the reference temperature. The background temperature is the surface temperature behind the infrared sensor.

[0035] For example, the following concepts can be used as reference temperatures: These can also be combined with each other: 1) background temperature measured by direct contact, for example in the form of a foil temperature sensor glued to the background; 2) Indirect, non-contact measurement of background temperature, for example in the form of an additional infrared sensor mounted on the back; 3) The infrared measurement signal may have large variations in time and / or position, but may be filtered through an evaluation zone before or after the evaluation zone currently being processed or multiple adjacent time and / or position evaluation zones. a) The data for the area is averaged or filtered in an appropriate manner, particularly to average out short-term and positional variations, but to reduce the variation in the reference temperature by a reference factor (1-A obj / A tot ) to reduce it. This reduces the obj / A tot will fluctuate less than the background temperature, especially when the relative object areal density within the stream of conveyed material is high. b) If statistical or image processing methods are used to select data points of the measurement signal in areas that are likely to represent a particularly low radiation component from the object surface and a large radiation component from the background, a reference temperature value can be determined from this using further statistical methods, the reference scaling factor of which is significantly closer to 1 than the averaging described in a), especially in the case of strands that run quietly against the background. 4) Indirectly measure the background temperature, for example by contacting the backside with a resistance thermocouple or thermocouple. During the heating and cooling cycles of the temperature control unit, non-ideal heat conduction in the background wall thickness creates a constant temperature gradient, causing the background temperature to lag behind the temperature measured on the outside, resulting in systematic measurement errors. This can be minimized using various methods and is particularly relevant for time-varying measurement backgrounds. a) Simulation of heat flow through the background wall and estimation of reference and background temperatures based on a model of the assumed heat flow and assumed or optimized thermal diffusion coefficient within the wall; b) additionally measuring the heat flow of a temperature control device behind the background using a heat flow sensor, and simulating the reference temperature and the background temperature based on a model of the measured heat flow and the assumed heat diffusion coefficient in the wall or the heat diffusion coefficient determined from the optimization; c) Symmetrization of heating and cooling cycles, where the rate of temperature change is equivalent when passing through minimum infrared contrast conditions. In this case, the FIFO data buffer always contains the same number of minima from heating and cooling cycles, and the regression model independently compensates for the fact that the reference temperature is sometimes slightly too high and sometimes slightly too low. 5) Furthermore, the redundancy that the reference temperature can be determined from both the measurement data from the contact or non-contact temperature sensor and the non-contact measurement data from the infrared sensor can be used for higher accuracy, better stability, mutual monitoring of the sensor systems, and joint calibration and analysis of the density of the material flow being conveyed. In particular, the reference temperature can be determined as follows: a) As a weighted arithmetic mean of concepts 3 and 4, b) By using the measurement data according to concept 3 to optimize the parameters of the heat conduction-based simulation model 4a or 4b so that the time and / or position offset is minimized. As a result, the reference temperature determined according to concept 3 differs from the reference temperature determined with optimized concept 4a or 4b by one scaling factor. A weighted arithmetic mean of concept 2 and optimized concept 4a or 4b can be used as the reference temperature for the regression. This includes the option of using only concept 3 or only one of optimized concepts 4a or 4b.

[0036] Variant 5b is particularly advantageous because, after the two independent reference temperature methods are reconciled, the same reference temperature is determined independently at the minimum of the contrast. In particular, after modeling according to concept 4a or 4b, the more reliable temperature measurement can determine the temperature compensation value Tc at regular intervals and adjust the measurement signal of the infrared sensor to the more reliable reference temperature (see FIG. 3 and the description).

[0037] Functional models that can be approximately determined from point clouds of noisy fluctuation data (e.g., regression and minimum squared error methods) or machine learning-based models are particularly well suited to accurately identifying the minimum fluctuation values. With such a functional model, the minimum value can be calculated directly using mathematical techniques. Advantageously, since the background of the measurement subject varies with time and / or position, it is obvious to set up a regression model that plots, for example, the amplitude of fluctuations against time or position. This is possible, but has the following drawbacks: the desired result, the temperature of the object, cannot be determined directly from the regression, but can be read off from the measurement signal at the time / position of the minimum determined in the second step. As measurement signals are known to be noisy, it is often helpful to apply appropriate data filtering methods. In this method, the recorded data of the measurement signal around the time or position of the detected fluctuation minimum are used for filtering, for example by averaging. Furthermore, a minimum of variation must be included in the recorded data, as simple extrapolation is not possible. Another drawback of direct regression versus time or position is that the functional relationship of the measured signal variation versus time or position can be distorted by nonlinearities in temperature change with time and / or position, so more complex, and therefore potentially unstable, modeling lends itself to a suitable functional approximation. One can also select data that is suitable for functional approximation of the minimum search. In particular, if the time-dependent variation of the background is realized by a temperature gradient with a constant rate of change over time, or by local variations due to a uniform temperature gradient field, one can calculate a simple direct regression model of the signal variation with time or position.

[0038] For regression modeling of variability, it is more advantageous to select a reference temperature, e.g., the background temperature, where the variability, e.g., the fluctuation amplitude, responds approximately linearly to the infrared contrast ratio of the object and background temperatures. The regression model defined above, with the reference temperature as the reference variable, allows the functionally determined position of the minimum variation to correspond directly to the determined object temperature. Each data point in the regression point cloud has two variations: along the longitudinal axis, the uncertainty in the determination of the degree of variation, and along the horizontal axis, the uncertainty in the determination of the reference temperature. The more data available, the more accurately the minimum can be determined. If the object temperature is stable, the redundantly collected data points will be close to each other. If the background temperature varies over time, several heating and cooling cycles can be represented in the point cloud, but with a correspondingly longer delay. If the background temperature starts lower than the object temperature and increases to approach the object temperature, the amplitude of the fluctuations decreases with increasing background temperature. The minimum of the fluctuation is reached at the minimum of the infrared contrast when the object temperature and the background temperature are the same, and as the background temperature continues to increase, the fluctuation amplitude increases again. This applies to the time curve of the heating cycle and the local curve with respect to the temperature gradient. The fluctuation magnitude plotted against the reference temperature forms a cloud of points with a minimum. In the region on the left, where the background temperature is low, the intensity of the fluctuation decreases with increasing background temperature, while in the region on the right, it increases.

[0039] If, for various reasons, you do not change the background temperature enough to reach and pass the minimum of the fluctuations, you can still determine the background temperature at which the minimum of the fluctuations occurs by extrapolation. The point cloud of the fluctuation intensity at the reference temperature does not include the minimum. Nevertheless, a regression analysis can be performed: the background temperature at which the fluctuation intensity extrapolated from the functional relationship reaches zero or a predeterminable value is output as the object temperature.

[0040] Area infrared cameras with gradient backgrounds should be used for processes that require highly dynamic measurements of rapid changes in object temperature. This arrangement ensures that the entire point cloud for the regression model is contained in a single image, allowing measurement evaluation of each exposure without delay.

[0041] If it is necessary to determine the change in object temperature with respect to position within the measurement area, which may be particularly important along the direction of flow of the conveyed material, multiple sections can be provided for this purpose, and the object temperature and exact measurement position can be determined in each section. Additionally, FIG. 9 illustrates how the temperature change over time and position can be used to determine the longitudinal temperature gradient.

[0042] Many infrared temperature measurement methods require the emissivity of the material being measured to be corrected using multiplicative emissivity correction or expensive two-color infrared measurement techniques. Plastics have high emissivity, typically 85-95%, which must be corrected to ensure accurate measurements. Determining the emissivity of each recipe and transferring it to the system requires significant laboratory work. An ideal black emitter has 100% emissivity and 0% reflectivity. The radiator temperature can be calculated directly from the infrared radiation using Boltzmann's law. For example, a plastic object with an emissivity of 90% will reflect 10% of its radiation as gloss at its surface. Since thread-like or granular objects in the stream of conveyed material are usually convex, the infrared sensor receives thermal radiation reflected from very different directions at different part surfaces of the object to be measured. At the edge of an object there is sometimes a set of specular reflection angles that reflect radiation emitted from adjacent objects of approximately the same temperature. This self-luminescence, which has the effect of increasing the degree of radiation, can probably still be corrected in the case of regularly arranged strands, but in the case of a randomly fluidized granule-carrying flow, the self-luminescence situation cannot be controlled.

[0043] As can be seen from the measurement device under mirror inclusion, the influence of emissivity can be almost eliminated. For this purpose, the temperature of the immediate surroundings of the measurement spot is controlled by a housing temperature control system so that the infrared radiation almost matches that of a blackbody radiator. Ideally, the exchange of radiation between the housing temperature control and the object to be measured is balanced so that all surfaces radiate the same amount of energy as they receive. For wear-resistant surfaces exposed to product wear dust, the housing temperature control cannot be realized as a black radiator. Edge effects due to limited installation space and the proportion of infrared light reflected by the housing temperature control must be taken into account. Ideally, the inner surface of the measurement background and the housing temperature control should be homogeneously diffuse-scattering, unpolarized, and have an emissivity of more than 85%, preferably at least 70%. If necessary, these properties can be achieved with special coatings. Advantageously, the surrounding surfaces of the temperature-controlled measurement background and / or the temperature-controlled guide body or the housing for guiding the flow of the material to be conveyed may be provided with a suitable coating, preferably made of a material such as plastic or plastic with high emissivity. Advantageously, the surfaces of said conductive bodies may be provided with a lacquer coating and / or a non-stick coating, for example made of fluoropolymers and / or silicone.

[0044] Since the measurement background already irradiates the inside of the housing temperature control with almost accurate infrared light, the influence of the irradiation error of the housing temperature control that deviates from the object temperature can be reduced. Radiative edge losses at the inlet and outlet regions of the enclosure temperature control are more problematic due to the limited installation length available for enclosure temperature control in most systems. Radiation losses also occur when the enclosure temperature control cannot be seamlessly adjacent to the temperature controlled background, such as in the case of a water bath with a background heating bar that crosses the strands. When all radiation losses are combined, the housing temperature control is illuminated from the inside with a slightly lower radiation intensity than would be the ideal radiation field for measurements under glare rejection. To compensate for the non-ideal emissivity within the housing temperature control and the radiation losses at the edge regions, the housing temperature control is adjusted to a temperature slightly higher than the measured object temperature.

[0045] To reduce edge effects and increase the effective emissivity of the conductor or the surface surrounding the measurement background, or bring it even closer to unity and thus save energy, the measurement background and / or said conductor, as well as the housing temperature control, can be equipped with thermal insulation. Alternatively or additionally, the temperature control device may include large heating and / or cooling elements on the surface of the temperature-controlled background or guide body, allowing the temperature of the surface to be efficiently and quickly controlled to reach the required target temperature in a rapid cycle. In particular, the planar heating and / or cooling elements on the temperature-controlled surface of the measurement background or guide body for the flow of the transported material in the measurement part area can manage a temporal and / or positional temperature gradient or the required background temperature change with time and / or position, during which the fluctuations in the measurement signal are analyzed in the described manner.

[0046] The device for measuring moving objects in front of a temperature-controlled measurement background usually needs to be adapted to the conditions of the production plant. The measurement background can be formed, for example, by a tubular or channel-like guide body through which the material to be conveyed is guided, but it can also be realized by a sliding surface, a side surface, a cover surface, a freely suspended object, a water-surrounded object, or an object immersed in water.

[0047] A device for changing the background temperature over time means that, for example, the background temperature of the measurement area of ​​the infrared sensor can be changed in heating and cooling cycles by a temperature control device, in particular depending on the object temperature. A pyrometer is preferred as an infrared sensor for this purpose.

[0048] Using an infrared line scan or area scan camera as the infrared sensor allows a larger measurement area to be analyzed against a background that can purely vary in temperature over time and can be adjusted uniformly.

[0049] However, in order to obtain measurement results more quickly, it is advantageous to have different background temperatures, especially when using infrared area scan cameras, and devices that locally change the background temperature are suitable for this purpose. A temperature variation with position means, for example, that the background is brought to different temperatures by a temperature control device along the distance covered by the flow of conveyed material, or that a temperature gradient is set up along the path of the conveyed material, for example, so that the background guide device through which the flow of conveyed material is conveyed is warmer upstream than downstream, or conversely, so that the upstream parts of the guide device are cooler than the downstream parts. By placing both mounting options directly next to each other, two evaluation sections can be realized, and from these two temperature measurements the temperature change of the conveyed material stream can be determined, which allows conclusions to be drawn about cooling measures, the core temperature of the object, or exothermic reactions. It should be noted that the infrared sensor may have multiple sensor elements distributed in the direction of the flow of the transported material, along the measurement section, or in the direction of the temperature gradient, in order to be able to capture variations in the sensor system signal in different parts of the measurement section.

[0050] The background temperature gradient can be at any angle to the flow of the material being conveyed, for example perpendicular. Non-uniform temperature fields with temperature gradients in different directions, circular temperature fields or periodic temperature fields are also possible.

[0051] The creation of a temperature gradient along the direction of flow can be achieved by heating and / or cooling elements of a temperature control device, which can be distributed along the background. Taking into account the heat conduction occurring in the background material, it may be sufficient to heat or cool only a short section of the guide device to provide two opposing temperature gradient fields for the two evaluation portions along the flow direction of the measurement section of interest.

[0052] The temperature control device may advantageously comprise one or more temperature sensors, by means of which the temperature of the measurement background or of the conductor or tubular body can be measured, said temperature sensors being able to operate in contact mode. The background temperature between the temperature sensors can be determined by interpolating the readings of two adjacent sensors at an appropriate distance. Alternatively or additionally, a non-contact sensor can be provided to determine or calibrate the background temperature field. Depending on the measured background temperature, the temperature control device can be controlled to generate a temperature gradient at the required position and / or time, which may include a temperature in its middle part that is approximately the same as the temperature of the object to be measured.

[0053] The measurement background can also be achieved by passing the infrared radiator through a window material that is sufficiently transparent to the infrared radiation. If transmission losses occur, the infrared radiator can operate at a higher temperature. In the following, the background temperature of such an emitter is defined as the sensor system weighted infrared radiance, which is evaluated to be the same as the normal radiation background at approximately ideal radiance and adjusted to this background temperature. To make the infrared radiation visible as the radiation background for different temperatures, the emitter can be calibrated with an appropriate characteristic curve.

[0054] A device for varying the background temperature over time and position to a target temperature, for example, has a temperature gradient field, and a temperature control device changes the average temperature across the temperature field with heating and cooling cycles. Such a device allows for the measurement of temperature gradients, particularly along the longitudinal direction of an object.

[0055] The control of the temperature control device, in particular the maximum and minimum temperature variations with respect to time and / or position, must be carried out automatically in conjunction with an evaluation device. Typically, a specific range around the object temperature is defined for the variations, which must follow the determined object temperature as symmetrically as possible. The temperature control device advantageously allows the average target temperature of the measurement background to be readjusted or readjusted, so that objects in the flow of conveyed material flowing through stand out brightly in front of cooler background areas and generate clearer fluctuation signals, while fluctuations in the sensor signal are minimized in thermally intermediate background areas, and objects settle in front of hot background areas with dimmer signals, so that stronger fluctuations in the infrared signal occur again.

[0056] In a further development of the invention, a tube can be used as the measurement background and guide body, which can be mounted rotatably about its longitudinal axis and can be rotated about its longitudinal axis manually or by a rotary drive, so that contrast measurements against the background are possible even with denser flows of the conveyed material, and for this purpose the tube can be placed in a position where it is rotated laterally. The infrared sensor aperture can also be offset laterally from the central axis, allowing the measurement spot to be aligned with the side of a nearby pipe. If the infrared sensor has a front mounting flange at an angle to the optical axis, the optical axis can be moved along the conical surface by simply rotating the sensor head flange. This allows the measurement spot to be aligned with different areas of the conveyed material stream, achieving relative object surface densities in the measurement area ranging from 15 to 80%, and more preferably 25 to 70%.

[0057] In an advantageous further development of the invention, a temperature control device can be used to keep the measuring head of the infrared sensor at a substantially constant temperature, so that the sensor system can operate within a narrow, predetermined temperature range despite fluctuating ambient conditions. In high and fluctuating ambient temperatures, this allows for improved measurement accuracy, improved long-term stability, and extended service life. Advantageously, such a temperature control device for an infrared sensor can include, for example, one or more temperature control elements on the sensor head of the infrared sensor. For example, a liquid temperature control unit, such as a water sleeve, can be provided on the sensor head to cool or heat the sensor head and thereby maintain it within a required temperature range.

[0058] To calibrate the infrared sensor, ideally, without interrupting power supply and temperature control, the sensor head can be relocated from the operating temperature measurement location to a nearby calibration station with a nearly ideal black emitter reference, which is very accurately and uniformly controlled to the relevant operating point temperature of the measurement spot location, and optionally the infrared scattered light field is also modeled based on the measurement situation, and the infrared sensor is calibrated to the operating point temperature at the relocated calibration location.

[0059] After the infrared sensor is repositioned, if there is no measurement object in the measurement field, a transfer calibration to a temperature-controlled measurement background can be performed immediately after the absolute operating point calibration, with the background ideally adjusted to the same operating temperature as the black emitter reference in a stable manner.A second parallel measurement is then performed using the infrared sensor repositioned to the measurement point, the background infrared radiation is measured simultaneously, and the background temperature is captured with a contact temperature sensor.From the results of the second parallel measurement, the irradiance of the temperature-controlled background environment can be determined and saved as a calibration parameter for this operating point in the absence of a flow of transported material.

[0060] To improve the accuracy of temperature measurement, the background and environment can be adjusted to the operating point temperature appropriate for the next production run when there is no measurement target, and the temperature compensation of the evaluation module can be updated using the calibration parameters for the operating point when there is no material flow. This allows targets to be measured directly with high absolute accuracy at the start of production, without the need for manual absolute calibration using a calibration station and black emitter standard. [Brief explanation of the drawings]

[0061] The present invention will be explained in more detail below with the aid of preferred embodiments and corresponding drawings, in which: [Figure 1] 1 is a schematic side view of an underwater pelletizing plant with an underwater pelletizer and a downstream centrifugal dryer, the outlet of which is equipped with a non-contact temperature measurement device for the emerging pellet stream. [Figure 2]FIG. 1 is a schematic side view of a dry-cut strand pelletizing installation with a water bath and two possible locations for a non-contact temperature measurement device on the strand and at the exit of the classification screen. [Figure 3] This is a perspective side view of the outlet and the non-contact temperature measuring device installed there, taken from the previous figure, including the basic scheme for measuring signal data processing, with the infrared sensor looking into the outlet pipe through the cutout on the side. This applies to all subsequent figures. [Figure 4] FIG. 4 is a perspective side view of an outlet similar to FIG. 3, in which the beam path of the infrared sensor is folded via a movable deflection mirror to allow a scanning motion within the outlet that is conveyed along with the flow of material being conveyed. [Figure 5] 1 shows a cross section of the outlet where the infrared sensor is mounted and the attached device for controlling the temperature of the measurement spot background and the tubular housing. [Figure 6] 6. Time curves of the measurement signal from an infrared sensor with a circular measurement spot (pyrometer) shown in FIGS. 3 to 5, in particular the measured background temperature and the measured fluctuation intensity and object temperature as they change over time. [Figure 7] An object moving in front of a temperature gradient background measured by an infrared sensor equipped with an area sensor (infrared camera) in Figures 3 to 5 is shown in terms of the temperature frequency distribution within the evaluation area. [Figure 8A] 1 is a perspective view of an infrared sensor with an area sensor (infrared camera) for measuring the temperature of a strand, where a background with, for example, a temperature gradient field is placed below the strand and the measurement area is optionally surrounded by a housing temperature control to minimize the effects of emissivity. [Figure 8B] A more complex design is shown in the background with two temperature gradient fields, which are used to simultaneously measure the object temperature in two parts along the direction of the flow of the conveyed material. [Figure 9] 9 shows the local variation of the measured signal variability from the infrared sensor of FIG. 8 when using an infrared camera, with a temperature gradient field in the background. [Figure 10] 1 is a perspective view of an infrared sensor on a strand, where a line or area infrared camera is used as the sensor, an infrared radiation background is placed below the strand, which may also be immersed in a water bath, and the measurement area is optionally surrounded by a housing temperature control to minimize the effects of emissivity. DETAILED DESCRIPTION OF THE INVENTION

[0062] As shown in the figure, non-contact temperature measurement can be used at various locations in large-scale plants for granule production. The sensor system is suitable for dry or wet cut strand pelletizing systems as well as underwater pelletizing systems. The sensor system can be used in these and other systems at a variety of measurement locations beyond those shown as examples.

[0063] Figure 1 shows a typical extrusion line equipped with an underwater pelletizer 12. A melt feeder 15 forces the polymer melt through a die plate 14 into a cutting chamber 16 of an underwater pellet dryer 13, where the discharged strand material is cut into granules underwater by rotating knives and transported by pipeline to a pellet dryer 17, where the granules and water mixture are separated. The separated granules are dried and discharged through an outlet 18. From there, the granules are typically sent to further processing stations such as classification screens, vibratory feeders, or heated vessels, where, for example, the granules undergo a crystallization or autocrystallization process.

[0064] The temperature of the pellets leaving the outlet 18 of the pellet dryer 17 must be within the process temperature range. If the granules are too cold, there may not be enough residual heat to evaporate any moisture remaining on the surface. If the granules are too hot, they may weld together and form agglomerates. The process temperature range is particularly narrow for polymers that are intended to retain a high residual heat so that there is enough energy remaining to initiate immediate crystallization. Some polymers can tolerate temperature deviations of only a few degrees Celsius, otherwise a reject product will be produced. In particular, non-contact temperature measurements can be made at said outlet 18, where the dried granules are discharged in the form of a conveying material stream 2. The granular matter in the conveying material stream 2 is usually not discharged in the form of a slow-flowing, compact granule stream, but rather travels or fluidizes through the tubular outlet 18 at more or less intervals. Warm, moist air from the pellet dryer 17 can also become entrained in the boundary layer of the granules. In normal operation, a slight negative pressure is created within the pellet dryer 17 by the suction fan, and a counterflow of drying air is drawn into the outlet 18 in the opposite direction to the conveyed material flow 2. Overall, complex flow conditions may prevail at the outlet 18 with different air flows at different temperatures being directed through the outlet 18 in addition to the particulate matter of the material stream 2 being conveyed.

[0065] An infrared sensor 1 can be assigned to said outlet 18 so as to be able to look into its interior and measure the temperature of the object 3 passing through.

[0066] The inner wall of the outlet 18 forms the background 4 for the temperature measurement, preferably in the case of a circular cylindrical outlet tube the inner surface of the tube forms said background 4 for the temperature measurement (see Figures 3 to 5).

[0067] FIG. 2 shows a conventional dry-cut pelletizer in which a melt feeder 15 presses a polymer or other material through a die plate 14, forming parallel strands 7 that are first passed through a water bath 8 before being cut by a pelletizer 12. Various further processing stations can be arranged downstream as explained in Figure 1. Here a classification screen 9 is shown, from which the conveyed material flow 2 is directed to an outlet pipe 18. For the temperature measurement of the strands 7, two positions of the sensor system are shown as examples. A first infrared sensor 1a is placed above the water bath 8 (see FIG. 10), a temperature-controlled background infrared radiator is placed in the water bath 8, and the strands 7 underwater are measured by the infrared sensor 1a. Another infrared sensor 1b, preferably with a temperature gradient background, is placed in the inlet area of ​​the pelletizer (see FIG. 8A). The thermal image of the strands 7 in front of the temperature gradient background allows for accurate and highly dynamic measurement of the individual temperatures of all strands 7, even when the strands are dynamically vibrating. The two-part temperature gradient background (see Figure 8B) can be used to check whether short-term strand temperature changes are occurring between the two evaluation parts along the direction of the flow of the conveyed material, for example due to heating of the strand skin by a hotter strand. An infrared sensor 1c at the outlet 3 can be used to measure the temperature of the cut granules in the fluidized stream 2 of conveyed material (see also Figures 1 and 3-5).

[0068] Various other strand pelletizing plants with automatic strand feeding with wet and dry cutting, not shown separately here, can also be equipped with non-contact temperature measurement technology for the strand and cut granules, and strand sensor systems can be used, especially in the inclined (chute) area. A small area of ​​the chute can be designed as a heated background, or background infrared radiators can be used in the water-cooled area.

[0069] As shown in Figure 3, the infrared sensor 1 can use its optical system to look inside the outlet 18 through the opening 19, so that the infrared sensor 1 has at least one measurement spot 20 within the outlet 18 through which the conveyed material flow 2 passes. When an infrared camera is used as the infrared sensor 1, a larger measurement area is captured and from each pixel or group of pixels a separate measurement spot 20 is measured. As an example, a moving granular object 3 is shown in the measurement spot 20 of the infrared sensor 1.

[0070] It is advantageous for the inner jacket surface of the outlet 18, and therefore the background 4, to be provided with a high emissivity, which can be achieved by a corresponding design of the outlet walls and / or by a suitable coating of the wall surfaces forming the area of ​​the measurement background 4 and the housing temperature control 22. For example, the inner walls of the outlet 18 can be coated with a plastic-like coating, a lacquer or a non-stick coating, especially made of fluoropolymers or silicones, which have a high emissivity.

[0071] The outlet 18 and thus the background 4 can be temperature controlled, at least in the region of the measurement spot 20, by a temperature control device 6, which can consist of at least one heating and / or cooling element 21 attached to the wall of the outlet 18.

[0072] Preferably, an enclosure temperature control system 22, for example in the form of a heating sleeve, can be provided near the temperature control device 6, which enclosure temperature control system 22 at least partially surrounds the outlet 18 and can have multiple regions for generating different temperatures in different parts of the outlet 18.

[0073] In particular, the temperature control device 6 may be configured to vary the temperature of the background 4 with respect to position and / or time during the measurement of the infrared sensor 1. The temperature of the entire outlet 18 can also be raised and lowered uniformly via the housing temperature control 22. If the temperature of the background 4 varies with position, the temperature control device 6 can heat and / or cool different background areas 4a-4c of the outlet 18 in different ways, for example to set up different temperature fields along the path of the conveyed material flow 2. For this purpose, the temperature field can have any shape, any number of regions, temperature minima and maxima, and temperature jumps, as long as different background temperatures occur simultaneously depending on the position. The temperature field can also vary with respect to time. Independent measurements of the background temperature 25 can be made using contact or non-contact temperature measurement techniques and can optionally be supported by a temperature simulation model that can incorporate measurement data from different positions on the background 4; measurement of the background temperature 25 is not required, but can still be provided. Since the conveyed material flow 2 usually does not completely cover the background 4, the reference temperature 38 can be determined only from the measurement signal 29, which is at least partially influenced by the background temperature 25. For example, in the case of a non-uniformly adjusted background 4, it is possible to start from any grid of the evaluation area 45 and determine the reference temperature for all areas individually. There is then the option of combining bins 45 of similar reference temperatures 38 into larger bins 45 and then using these larger bins 45 to determine their variability 32 and reference temperatures 38 .

[0074] The temperature of the background 4 may be controlled by a temperature controller 6, and one or more temperature sensors 23 may measure the temperature at the outlet 18 and control the temperature controller 6 accordingly.

[0075] Data processing of the measurement signal 29 takes place in the evaluation device 30. In the case of several evaluation portions 51 (see FIG. 8B or FIG. 9), a respective subset of the measurement data is assigned to each portion 51 in a first pre-processing step.

[0076] In a second pre-processing step, the measurement signal 29 is optionally corrected with a temperature compensation value Tc 40 .

[0077] One complete data processing operation is then carried out for each portion 51 and the determined object temperature 39 is output in each case.

[0078] The individual data processing procedures are as follows. The fluctuation intensity of the measurement signal 29 is evaluated by a fluctuation evaluation module 31 over a certain amount of signal data including a time and / or position evaluation zone 45 of similar background temperature and output as a fluctuation degree 32, which is essentially proportional to the measured infrared contrast between the object and the background, e.g., in the form of a fluctuation amplitude. Alternative methods for determining the fluctuation degree 32 are presented herein. Additional sets of signal data from other evaluation zones 45 are similarly evaluated, with different variability 32 available for different background temperatures 25 . If the background temperature 25 varies over time, a certain amount of time is required to capture different evaluation zones 45, and if the background temperature 25 varies over location, different evaluation zones 45 can be captured and evaluated simultaneously. For each time and / or location evaluation zone 45, a reference temperature 38 is determined in a reference temperature determination module 37, which can be used as a reference variable in the regression model 33. The reference temperature 38 can be calculated, for example, from the average background temperature 36 or the average infrared measurement signal 35, a weighted arithmetic mean thereof, and various other methods. In particular, the description shows how image processing, statistics and heat diffusion modelling can be used to internally determine two reference temperatures using two independent calculation paths: modelling a - based on infrared measurement data 29, and modelling b - based on background temperature 25 and optionally heat flux measurements 54 from a heat flux sensor 53. From the difference Δ (delta) between modelling a and modelling b, in a pre-processing step in the evaluation device 30, the temperature difference Tc 40 for the minimum fluctuations can be determined at regular intervals for the aforementioned temperature compensation of the measurement signal 29.

[0079] In the regression module 33, the functional relationship of the variability 32 to the reference temperature 38 is determined, and the determined object temperature 39 is determined by a minimum search 34 or by extrapolation of the intersection of the regression function with the zero axis or a predetermined value.

[0080] If the background 4 is divided into multiple evaluation portions 51 (see Figures 8B and 9), which can be as small as a pixel line, the object temperature 39 can be determined for each portion using a section-specific regression model 33 for each section area. For example, if the background temperature field has a complex temperature gradient 48 (see, e.g., Figure 8B) and minima in the infrared contrast of the object 3 against the background 4 occur in two or more evaluation portions along the path of the flow of the conveyed material, a certain position resolution can be reached in the direction of the flow of the conveyed material 2 by determining the local object temperature 39 in locally different portions 51. The more areas 51 with properly adjusted background 4 are provided, the finer the spatial resolution.

[0081] For example, in the evaluation section, the upstream portion 4a of the outlet 18 can be at a lower temperature than the conveyed material stream 2. The central portion 4b of the outlet 18 can be adjusted to at least approximately the same temperature as the conveyed material stream 2. The downstream portion 4c of the outlet 18 can be at a higher temperature than the conveyed material stream 2 (see FIG. 7).

[0082] FIG. 4 shows a periodically deflected measurement spot compared to a fixed measurement spot as shown in FIG. The beam path of the infrared sensor 1 is bent at a dynamically changing angle, for example via a rotating prism mirror 24. The measurement spot 20 or the measurement area of ​​the infrared camera is moved by the deflection in order to follow the direction of the conveyed material flow 2 at as close a speed as possible to that of the moving object 3. For this purpose, an opening 19 is machined in the form of an elongated slit at the outlet 18 in the direction of movement. The temperature control device 6 or its heating and / or cooling element 21, as well as the housing temperature control 22, are also designed to be correspondingly long so as to be able to control the temperature of a longer outlet section or background section.

[0083] If the rotation speed of the prism mirror 24 is continuous, the scanning speed of the measurement spot 20 will not be exactly constant due to changes in the angular relationship or distance. However, since the rotation speed of the prism mirror 24 can be controlled according to the angle, the measurement spot 20 actually moves along the background 4 at a constant speed. If the scanning speed of the measurement spot 20 is well matched to the speed of the conveyed material stream 2, the time that the infrared sensor remains stably aligned with the moving object 3 is significantly increased. This allows the temperature of each individual moving object 3 to be measured and the temperature distribution of objects within the conveyed material stream 2 to be determined. The temperature control device 6 advantageously consists of a number of individually controlled heating / cooling elements 21, so that a temperature gradient field in the background 4 is formed, in particular along the lateral or scanning direction. The background temperature at the position where the contrast of the moving object 3 disappears indicates the temperature of this individual object.

[0084] 5, the infrared sensor 1 can also be used to view the inside of the outlet 18 at an angle. When the measurement point 20 is fixed, such an oblique arrangement can be made at an angle of, for example, 30° to 80° or 35° to 55° with respect to the longitudinal axis of the outlet, but in principle this is also true when the above-mentioned prism mirror 24 is used. The housing temperature control 22 is designed here in the form of a heated sleeve with a recess for the temperature control device 6. A water temperature control sleeve 5 is arranged around the measuring head of the infrared sensor 1, which allows temperature-stabilized operation of the sensitive sensor system. Advantageously, the infrared sensor 1 can be operated over a narrow temperature range with active temperature control, allowing the sensor system 1 to measure accurately with maximum precision over long periods of time despite fluctuating ambient conditions, and this temperature control can preferably be achieved by placing a water temperature control sleeve 5 around the sensor head.

[0085] To protect the lens of the infrared sensor 1 from dust and dirt, finely filtered instrument air flows downward from purge air nozzle 27 towards the conveyed material stream 2 . A combined temperature and heat flow sensor 53 may optionally be used at the heat transfer interface from the temperature control device 6 to the outlet 18 to measure the heat flow transferred to the outlet 18 during heating and cooling operations of the temperature control device 6, as described herein, and more accurately determine the temperature of the background 4 from the thermal conductivity coefficient seen through the wall material.

[0086] As shown in Figure 6, the background temperature 25 of the background 4 varies over time, as is possible with a temperature control device 6 according to any of Figures 3 to 5 or 10. The infrared measurement signal 29, shown here in simplified form as a thin wavy line, varies very dynamically, since at the measurement spot 20 of the infrared sensor 1, sometimes more infrared radiation is received from the object 3 and sometimes more from the background 4. The fluctuation amplitude 32a of the measurement signal 29 increases the further the background temperature 25 is from the object temperature 26. The measurement signal 29 practically always fluctuates only between the object temperature 26 and the background temperature 25, which means that this fluctuation is almost non-existent when the object temperature 26 and the background temperature 25 are the same, i.e. when the radiation from the object becomes indistinguishable from the radiation from the background 4 in the spectral evaluation of the infrared sensor 1. In the situation of such a contrast minimum 42, the determined object temperature 39 value can be obtained directly from the measurement signal 29, from the measurement signal 35 averaged, for example using a low-pass filter, or from the background temperature 25.

[0087] In order to pass through such a contrast minimum 42 regularly, the background temperature 25 can in particular be periodically increased and then decreased again, with the mean value between the increase and decrease in temperature also being changed at the same time, in particular so that the mean value of the background temperature 25 approaches the object temperature 26. As Figure 6 shows, initially the temperature cycle is too low, which means that the boost cycle is still below the object temperature 26. However, if the average temperature is also adjusted, the background temperature 25 and its variation can be set so that the background temperature 25 oscillates around the object temperature 26 (see the right half of Figure 6).

[0088] In particular, the degrees of temperature variation of the background 4 are selected so that as the background temperature 25 changes, these variations 32 change, with the variations 32 periodically passing through a minimum value as the infrared contrast between the object and the background disappears. The variability 32 is plotted on the right vertical axis of Figure 6, where "αSD[TIR]" is proportional to the standard deviation of the infrared measurement signal. As the background temperature 25 increases above the object temperature 26, the variability 32 increases. When the increase in background temperature 25 is again reduced, the variability 32 decreases again until it reaches a minimum value of variability. Thereafter, when the background temperature 25 becomes lower than the object temperature 26, the variability 32 increases again, and when the background temperature 25 rises again from the decreased state, the variability 32 reverses, and when the background temperature 25 reaches, for example, the object temperature 26 again, the minimum value of the amplitude occurs again.

[0089] The measurement signals 29 from the infrared sensor 1 are evaluated by an evaluation device 30 as explained in Fig. 3. In evaluation zones limited by time 45, which are short enough to assume that the background temperature 25 remains almost unchanged, the degree of variability 32 for each evaluation zone 45 is calculated. The acquisition period during which the measurement signal 29 is recorded to enable the degree of variability 32 of the evaluation zone 45 to be determined can be defined within a wide range, consisting of only a single data point, up to the point at which the actual variability analysis is carried out in a subsequent regression calculation. However, it is often advantageous to reduce the large amount of measurement signal data 29 to a few characteristic values ​​at an early stage and summarize the data, for example, in the range of a tenth of a second to a few seconds. Various calculation options are presented herein for calculating the variability 32. Individual measurements of the measurement signal 29 are recorded over short periods in order to determine the variability 32 of the measurement signal 29.

[0090] In parallel, measurements of the background temperature 25 are averaged in the same cycle. In this example, the average background temperature 25 associated with the time-synchronous data of the variability 32 is temporarily stored as a reference temperature 38, for example in a FIFO data buffer, for regression evaluation.

[0091] If measurement data from one or more temperature sensors 23 are available for the background 4 independently of the infrared sensor 1, the contrast minimum 42 provides a good opportunity to co-calibrate the measurement signal 29, which may be affected by many disturbance variables, to the background temperature 25, which can be determined more reliably, for example via the same temperature sensor 23. Furthermore, during a phase in which the object temperature 26 remains apparently constant over a long period of time, the background temperature 25 can be controlled as close as possible to the object temperature 26 for a certain period of time. In this minimum contrast steady state, there is sufficient time for the temperature sensor signal 23 to adjust to the background temperature 25, and in particular, if it is confirmed that a temperature change in the background temperature allows for the determination of an almost identical object temperature 39 before and after the steady state phase, the difference between the infrared measurement signal 29 measured by the temperature sensor 23 and the background temperature 25 can be directly used as the temperature compensation value 40.

[0092] Figure 7 shows the evaluation of a measurement signal image or part thereof as a cross-sectional image of a multi-sectional measurement signal image of an infrared sensor 1 in the form of an area infrared camera, such as can be used in the device according to figures 3 to 5. Figure 7 shows, in a pseudo-colour representation for visualising the temperature, a moving object 3 in a stream 2 of conveyed material bending in the X direction 41 against a background 4 with a uniform temperature gradient field 48. The background 4 is cooler than the object 3 in the upstream part 4a, close to the object temperature 26 in the middle part 4b, and warmer than the object 3 in the downstream part 4c. The temperature field of the background 4 can also have a rather complex non-uniform temperature field with different gradient directions and different temperature zones and temperature jumps, which creates the prerequisites for measuring the object 3 moving at different locations relative to the background temperature 25. The evaluation zone 45 constituting a particular background temperature range can be realized in a uniform temperature gradient field in a simple manner by a rectangle approximately bounded by isotherms 44. In the case of a non-uniform background temperature field, for example, more complex shapes or grids can be used to bound the evaluation zone 45 in terms of location (see herein).

[0093] Due to the short residence time of the object 3 at the measurement spot 20 of pixel 47 relative to the response time of the infrared sensor 1, the measurement signal 29 at the excited pixel 47 reacts with a time delay, producing an afterglow trace with soft edge transitions, which is visualized here in simplified form as an ellipse 3 with sharp contrast. 4, this afterglow trace can be minimized, so that particles moving exactly at the scan speed of the moving measurement spot are mostly imaged at their true object contour, while for individual objects 3 moving slightly slower or slightly faster than the scan speed, only short afterglow traces are formed towards the front or rear. This means that the dwell time of the object 3 in front of the controlled pixel 47 is long enough for the measurement signal 29 to reach a maximum level, from which, using simple image processing, the individual object temperature 39 can be determined for each individually imaged object 3.

[0094] To determine the average object temperature 39, different degrees of variation 32 in different evaluation zones 45 are determined from the infrared image of the measurement section, in the example shown the evaluation zones are formed by temperature ranges of + / - 0.15°C around the respective average background temperature 25, with the isotherms 44 as limits. By way of example, nine evaluation zones 45 are evaluated in Figure 7. For this purpose, the measurement signals 29 of all pixels 47 within each evaluation zone 45 are statistically evaluated, the standard deviation being particularly suitable for determining the degree of variability 32. Alternative statistical parameters such as range and interdecile distance are described herein. To visualize the statistical distribution of the measurement signals 29 within the rectangular evaluation zones 45, a temperature histogram 43 is shown at the bottom of Fig. 7, which refers to the data for each evaluation zone 45 in the infrared image above, with only the evaluation zone 45 being explicitly drawn with reference to histogram 43b. A reference temperature 38 for each evaluation zone 45 is determined, which can be done, for example, by averaging the background temperature 25 of this area or by other methods as described.

[0095] In the false color image, the local infrared contrast relative to the central axis of each evaluation area corresponds to the relative temperature T rel 46 is shown by the black dashed line, and T rel is calculated as the difference between the measurement signal 29 and the reference temperature 38 for each evaluation zone 45 . The infrared contrast can be seen in the distance between the maxima 49 and 50 of the temperature distribution of the temperature histogram 43. All histograms 43a-i have the same axis scale for temperature, and in each case only the reference temperature 38 is labeled. The distance between maxima 49 and 50 is particularly large in the downstream portion 4c and the upstream portion 4a of histogram 43i. In the middle portion 4b, temperature histogram 43e shows only the central maximum, since the object temperature 26 of object 3 is approximately the same as the background temperature 25 without infrared contrast. In this intermediate portion 4b, the pixels 47 of the infrared sensor 1 are approximately stable with their associated measurement spots 20 in balanced radiation exchange, so that instead of the regression method, the determined object temperature 39 can also be read directly from the histogram 43e with the minimum standard deviation as the temperature value of the main maximum 49, in this example 81°C. The disadvantage of this method is that during successive measurements at different positions of the object 3, the determined object temperature 39 may jump from measurement to measurement between different evaluation zones 45.

[0096] In the evaluation method, the angular orientation of the transported material flow 2 relative to the background temperature gradient field 48 is irrelevant if it can be assumed that the object temperature 26 is approximately constant at all locations within the infrared image or at parts of the infrared image. If the object temperature 26 undergoes a non-negligible cooling while moving through different areas of the background 4, it may be advantageous to orient the temperature gradient 48 of the background 4 opposite to the direction of the transported material flow 2. This causes the spatially varying object temperature 26 to clearly intersect with the temperature of the temperature gradient field 48 and not run parallel in some areas, which can cause problems in the minimum search 34 in areas of constant variation 32. If the moving object 3 may be heated during the measurement due to an exothermic process such as a chemical reaction or crystallization, it may be advantageous to arrange the temperature gradient field 48 parallel to the flow 2 of the conveyed material.

[0097] If the object temperature 26 is expected to change in an unknown direction during movement across the background 4, a conveyed material stream 2a that curves transversely to the direction of the temperature gradient field 48 can be used, and a uniform object temperature 26 transversely to the conveyed material stream 2 can be assumed. In most cases, it is sufficient to provide a sufficiently steep temperature gradient in the background 4 relative to the gradient caused by changing the object temperature 26 as the object 3 moves across the background 4 .

[0098] Figure 7 also implies a special situation of a background 4, i.e., very narrow, e.g., with overall dimensions of only the width and height of a single white-filled evaluation area 45 above histogram 43b. As explained in the introduction, this narrow background 4 can have any temperature gradient field 48, the width of which must, however, be assumed to be narrow enough that the temperature gradient in the X direction is negligible. Regarding the curve of the background temperature 25 in the Y direction, we will discuss two relevant scenarios in particular. Scenario A: In this special situation, the entire background 4 of the infrared sensor 1, or the white-shaded image area constituting the entire section of the image, has a temperature gradient oriented in the Y direction. This scenario is completely covered by the explanation just given, since it is nothing more than a 90° rotation of Figure 7, with the only difference being that the conveyed material stream 2a is bent vertically and the afterglow trace of the moving object 3 is formed accordingly in the V direction. Scenario B: In this special situation, the white shade image area that constitutes the entire background 4 of the infrared sensor 1 or the entire section of the image is adjusted uniformly in the V direction.

[0099] The associated histogram 43b visualizes the statistical distribution of the measurement signal 29, from which a variability 32 that is approximately linear with the infrared contrast can be determined, for example, using the standard deviation method. Other statistical methods are described in the specification. The infrared contrast situation currently visualized in temperature histogram 43b is as follows: The background 4 is limited in this particular situation to only a white shade evaluation area 45 and is uniformly conditioned to 78° C. The relative object surface density in the conveyed material stream 2 is fairly low and the maximum value 49 is dominated by pixels 47 which receive substantially all infrared radiation from the surface of the background 4.

[0100] Because all pixels 47 are periodically driven by the warmer object 3 to higher measurement signal values ​​29 closer to the object temperature 26 of 81°C, the dominant maximum 49 of temperature histogram 43b is slightly higher than 78°C due to the afterglow effect. Similarly, secondary maximum 50 does not fully reach object temperature 26 and is slightly lower than 81°C because pixels 47 are not exposed to the stronger infrared radiation from object 3 for long enough. This is better achieved with an apparatus such as that shown in FIG. 4, where maxima 49 and 50 can reach much closer to the exact values ​​of background temperature 25 and object temperature 26, as explained in detail elsewhere. In summary, the evaluation of the measurement signal data 5 visualized in the temperature histogram 43b for this special situation of the background 4 with the size of the white-shaded evaluation area 45 can be explained in such a way that the background temperature 25 is still significantly lower than the object temperature 26, and then in order to further increase the background temperature 25, it is necessary to capture the variability 32, which at least initially decreases, and determine the object temperature 39 by regression analysis from the data collection of the different variability 32 over time at different background temperatures 25. This method is the same as that described in Fig. 6 for varying the background temperature with respect to time and processing the degree of variation 32 with a regression model 33. Therefore, for the analysis of the variations of the measurement signal 29, it is irrelevant whether these are determined in time via the variations of a single pixel 47, i.e. a single measurement spot 20 of the infrared sensor 1 designed as a pyrometer, or in position via the variations of the infrared intensity captured in the line of the infrared sensor 1 through the narrow background strip described as a special situation. Instead of an infrared line scan camera, an infrared area scan camera can also be used as the infrared sensor 1. In this case, only the pixels 47 of a measurement section in the form of a strip-shaped image section (region of interest, ROI) are read, whose measurement spot 20 is aligned on a calibrated strip-shaped background 4, as can be realized with the device of FIG.

[0101] FIG. 8A shows a version of an infrared sensor 1 in the form of an area infrared camera aligned within the field of view of one or more strands 7 moving in the longitudinal direction of the strands as the stream 2 of conveyed material moves. To reduce the influence of the emissivity of the strand 7 being measured, an enclosure temperature control 22 is recommended. To reduce edge effects, the opening 19 should not be larger than necessary. If possible, the specular reflection from the strand surface photographed from the camera's point of view should come from a spatial direction covered by the enclosure temperature control 22, but certain edge effects due to infrared radiation at the inlet and outlet ambient temperatures 55 can hardly be prevented. Behind the strands 7 there is a measurement background 4, which has some structured temperature field with the help of the temperature control device 6, so that different background temperatures 25 are simultaneously present at different locations in the field of view of the infrared camera 1. Additional temperature information 25 of the background 4 can be determined by a contact or non-contact temperature sensor 23, so that the background temperature 25 at each location of the background 4 is at least approximately known from this information and the information of the infrared sensor 1. However, it is also sufficient to determine a reference temperature for a temperature range with a similar background temperature 25, as explained in the data processing scheme of Fig. 3. To be able to measure the strand temperatures 26 individually, it is advantageous if the temperature gradient field 48 of the background 4 has a temperature gradient that is oriented substantially in the direction of movement of the strands 7. For example, if a thin layer of water is still attached to the strand 7, the strand temperature 26 will decrease along the direction of the transport material flow 2 due to a strong ablation cooling effect. Therefore, to determine the string temperature with a stable regression model 33, it is desirable to position the temperature gradient field 48 of the background 4 in the opposite direction to the temperature gradient of the string 7, i.e., warmer than the strand 7 in the downstream portion 4c and cooler than the strand 7 in the upstream portion 4a. The temperature gradient field 48 directed in the opposite direction should have a temperature gradient that is sufficiently steeper than the gradient of the strand temperature 26 in the longitudinal direction of the strand.

[0102] Further sections with other background temperature gradients 48 can also be used in the direction of the conveyed material flow 2 to measure the temperature gradient of the strand temperature 26 in the longitudinal direction of the strand. The regression models 33 assigned to the individual evaluation sections 51 determine the local object temperatures 39 of each of the individual sections.

[0103] FIG. 8B shows such a simple subdivision of the background 4 into portions 51, where the object temperatures 39a, 39b are measured independently of each other for two slightly overlapping evaluation portions 51a and 51b at different positions 41a, 41b in the flow 2 of the transported material. The temperature control device 6 is used to set a low background temperature 25 in the upstream section 4a, a background temperature 25 higher than the strand temperature 26 in the middle section 4b, and an even lower background temperature 25 in the downstream section 4c than in section 4a. In the evaluation section 51a, the temperature gradient field 48a is curved relative to the direction of the transported material flow 2, while in the evaluation section 51b, a background with a stronger temperature gradient field 48b is carried out along the direction of the transported material flow, which means that separate evaluations can be made for both evaluation sections 51a and 51b. As an example, strand 7 is shown in a temperature visualization pseudocolor display against the background, with the strand temperature 26 decreasing along the direction of the conveyed material flow 2. In evaluation section 51a, a fluctuation minimum is reached in the evaluation zone at position 41a, thus a target temperature 39a of approximately 80°C. Another temperature evaluation in a second evaluation section 51b downstream evaluates a further fluctuation minimum at position 41b, where a target temperature 39b of approximately 76°C is measured. The longitudinal temperature gradient of strand 7 can be simply determined as the differential temperature between 39a and 39b divided by the distance 52 between measurement positions 41b and 41a, the exact positional value of which is determined from the regression of each evaluation section.

[0104] FIG. 9 shows a method for evaluating the measurement signal image of an infrared sensor 1 in the form of an area infrared camera that can be used in the device according to FIG. 8A. This false color image visualizing the temperature can also represent evaluation portions, i.e. image sections, so that multiple evaluations are calculated section by section for the entire measurement signal image (see Figure 8B). The only difference to Figure 7 is that here, instead of moving strand-like or granular objects 3, continuous strands 7 run against the background 4. The flow 2 of the conveyed material is essentially achieved by the continuous longitudinal movement of the strands 7 in the X direction. Depending on the measurement position and guide of the strand 7, the strand 7 moves slightly in the Y direction but can exhibit dynamic vibrations, especially in the feed area of ​​the pelletizer. In production plants, it is often necessary to measure several dozen parallel wires 7. The background 4 has a temperature similar to the object temperature 26 and the temperature 25 varies depending on the position; in this case, the background 4 with its temperature gradient field 48 allows the infrared sensor 1 to be positioned relatively far away so that the strand 7 can be detected approximately with the measurement resolution of the infrared camera 1.

[0105] For fluctuation assessment, it is sufficient that the strand width is imaged by only one or two pixels 47, and that all partially covered pixels 47 receive thermal radiation from both the strand 7 and the background 4, even when the strand 7 is vibrating.

[0106] Due to the low strand speed, the strand is usually cooled along the direction of the conveyed material flow 2. In this application in particular, the information on the design of the gradient field detailed in Figure 7 must be observed.

[0107] To meet the requirement of measuring the temperature of each strand 7 separately at a specific X position, the background 4 can be divided into individually temperature-controlled parts in the Y direction. This makes it possible to simultaneously set the minimum value of the infrared contrast for all strands 7 in all parts very close to a specific X position, despite temperature gradients along the strands 7 and different average strand temperatures in the intermediate parts 4b.

[0108] It is also possible to define small portions 51a and 51b in the image in order to be able to determine in particular the temperature of individual strands 7. As the position of the strands 7 is not always fixed, it is recommended to use conventional image processing, in particular threshold segmentation, to define portions 51 around each strand or group of strands in the infrared image. The width of the portions 51 is preferably based on the width of each of the strands 7, thereby maintaining a stable relative object area density in each area of ​​each portion 51 for variability assessment.

[0109] If the strand temperature 26 is sufficiently stable in time, there is an easier way to determine the exact temperature of all individual strands 7 at a particular X-position. For this purpose, the average background temperature 25 of the background 4 with the temperature gradient field 48 is periodically increased and decreased, and a narrow evaluation section 51 following the strand is defined for each strand 7 with an appropriately selected background area section. Because the background temperature 25 varies with time and position, the X position of the contrast minimum shifts along with the average background temperature 25 across all sections 51, and the object temperatures 39 determined for each strand can be used at different X positions. Regression can be used to determine the longitudinal temperature gradient for each line 7. The respective measurement result of the determined strand temperature 39 at the target measurement position X is calculated from the currently determined object temperature 39, the measurement position 41 with the smallest variation, and corrected by the product of the difference between the target position and the measurement position and the respective longitudinal temperature gradient of the strand 7.

[0110] Figure 10 shows an infrared sensor 1 on a strand 7, similar to Figure 8A, and also includes optional enclosure temperature control 22. However, the environmental conditions here are more challenging, as the background 4 is immersed in a water bath 8, or at least operating in a water bath, or being sprayed with water. Direct water contact with the heated background 4 works reliably long term only in very limited system configurations. Deposits and flaking deposits cause uncontrollable radiation from the heated surface, making reliable infrared radiation from the background impossible.

[0111] A detailed view of the protective enclosure 10, specially developed for contact with water, shows that the temperature control device 6 is thermally isolated from the surrounding water in the form of a temperature control beam. A centrally mounted heating and / or cooling element 21 within the protective enclosure 10 is controlled to a background temperature 25 by a temperature sensor 23 and emits light through an infrared-transparent window 11 towards the surface of the background 4 and the infrared sensor 1. Infrared-transparent polymer films are known; see, for example, Garrett Beals, Gregory Baronek, Corey Smeaton, and Joseph Sperry, "Characterization of thin polymers for infrared windows," Proceedings of the International Society for Optical Engineering (SPIE) 12103, Advanced Optics for Imaging Applications, Ultraviolet (UV) to Long-Wave Infrared (LWIR) VII, 1210309 (May 27, 2022), https: / / doi.org / 10.1117 / 12.2618378. Such infrared-transparent polymer films are described herein. These can be waterproofed and attached to appropriately designed protective enclosures. Since the infrared-transparent window 11 absorbs part of the radiation emitted from the background 4, the temperature control device 6 placed inside can operate at a higher temperature and simulate the background temperature 25 according to the intensity evaluation of the infrared sensor 1. To obtain an accurate measurement, the infrared sensor 1 can be calibrated using a calibration radiator in an equivalent measurement arrangement. The temperature control of the temperature control device 6 can then be determined using transfer calibration.

[0112] The temperature control device 6 can be used to set a strand-specific background temperature radiation 25 by means of a number of heating and / or cooling elements 21 transverse to the strand 7, but this effort is of little relevance for practical applications. As a background radiation field, it is easier to realize a temperature gradient field comparable to that of FIG. 9. The statistical contrast can already be determined with a single heating and / or cooling element 21, which can be characterized by a temperature frequency distribution, such as the row area in Figure 9. This frequency distribution can be used to determine whether the radiative emission from the background 4 is higher or lower than the radiative emission from the strands 7.

Claims

1. 1. A method for non-contact temperature measurement of filamentous and / or granular objects in a conveyed material flow (2), comprising: an infrared sensor (1) is directed at the flow of material (2) flowing against a background (4), and the temperature (26) of the filamentous or granular object (3) is determined from a measurement signal (29) from the infrared sensor (1); The temperature (25) of the background (4) is varied in time and / or position by a temperature control device (6), The evaluation device (30) determines the degree of fluctuation (32) from the measurement signal (29) from the infrared sensor (1) at different background temperatures (25), determines the position of the minimum fluctuation from these fluctuations, and determines the object temperature (39) from the value of the measurement signal (29) at the time / position of the minimum fluctuation. A method characterized by:

2. the measurement signal (29) is subjected to a regression analysis by a regression analysis module (33) of the evaluation device (30), and a functional relationship between the change in the fluctuation degree (32), in particular the change in the fluctuation amplitude, and the background temperature (25) varying over time and / or position is determined by the regression analysis module (33); the time / position at which the fluctuation minimum, in particular the amplitude minimum, occurs is determined by the evaluation device (30) using the determined functional relationship, 2. The method of claim 1, wherein the object temperature (39) is determined from the value of the measurement signal (29) at the time / position of a fluctuation minimum determined from the functional relationship.

3. the background temperature (25) and / or the average measurement signal (29) and / or the reference temperature (38) are used as reference variables for the functional relationship of the fluctuation degree (32), in particular the fluctuation amplitude, 3. The method of claim 2, wherein the values ​​correspond as closely as possible to the values ​​of the background temperature (25) and the values ​​of the infrared measurement signal (29) in the temporal and / or positional regions where the degree of variability (32) is minimal, and in other temporal and / or positional regions, have a linear gradient proportional to the difference between the background temperature (25) and the object temperature (26) and a curve as linear as possible with respect to the background temperature (25).

4. The reference temperature (38) was used as the reference variable for the regression; the degree of fluctuation (32) of the reference temperature (38) is determined and / or taken into account only above or only below the object temperature (26), so that the degree of fluctuation (32) does not reach a minimum value; the determined object temperature (39) is determined in an extrapolation from a regression model (33); 4. The method according to claim 2 or 3, wherein the degree of variability (32), in particular the amplitude of variability, approaches zero or becomes smaller than a predetermined minimum variability threshold.

5. In a further evaluation step, the time and / or location of said minimum fluctuation is determined, in particular using a data record comprising the fluctuation degree (32), the reference temperature (38) and the corresponding time and / or location in different temporal and / or location evaluation zones (45), Contiguous data groups around the minimum variation value having a variability (32) less than a predetermined threshold are filtered; For each data group, a regression of the time or the location against the reference temperature (38) is performed; 5. The method according to claim 2, wherein for each data group the determined object temperature (39) is inserted into the respectively determined regression equation, thereby determining the time and / or location of minimum fluctuations.

6. The background temperature (25) varies with heating and cooling cycles; The heating half-cycle begins with the background temperature (25) being lower than the object temperature (26), reaches the object temperature, and then exceeds the object temperature; This is followed by a subsequent cooling half-cycle in which the high background temperature (25) is initially reduced, and then the object temperature (26) is reached again over a further period of time in order to return to the initial temperature below the object temperature; Meanwhile, the calculated fluctuation intensities (32) and the average background temperatures (36) or the average measurement signal values ​​(35) or the reference temperatures (38) for the same zones are recorded quasi-continuously in a time and / or location evaluation zone (45) limited to the temperature range of the background temperatures (25) and / or the average measurement signal values ​​(35) in a data memory, in particular a FIFO memory, 6. The method according to claim 2, wherein one to two minimum values ​​of the fluctuation intensity (32) are further processed from the recorded data, and with each new data set recorded the oldest recorded data set is no longer used for the evaluation, and when using the described regression model, an updated determined object temperature (39) can be output for each measurement cycle.

7. The background temperature (25) is varied by a temperature control device (6) through heating and cooling cycles, without reaching the object temperature (26); Meanwhile, the calculated fluctuation intensity (32) and the average background temperature (36) or the average measurement signal value (35) or the reference temperature (38) are recorded in a data memory, in particular a FIFO memory, so as to vary quasi-continuously in a time and / or positional evaluation zone (45) limited to the temperature range of the background temperature (25) and / or the average measurement signal value (35), for the same zone; 6. The method according to any of claims 2 to 5, wherein values ​​of the fluctuation intensity (32) of the minimum and maximum background temperature (25) in the transition between the heating cycle and the cooling cycle are further processed from the recorded data to output an updated determined object temperature (39) for each measurement cycle, using the described regression model used for extrapolation.

8. 8. The method according to claim 1, wherein the thermal radiation field of the background (4) is varied with respect to time and / or position by the temperature control device (6), and at a specific time and / or at least one specific position, the thermal image of the background (4) and the material flow (2) passing through the background provided by the infrared sensor (1) is evaluated by an evaluation unit (30) as being at least substantially contrast-free, and the object temperature (39) is determined from the position or time evaluated as being contrast-free.

9. The method according to any one of claims 1 to 8, wherein the temperature of the background (4) is controlled by the temperature control device (6) so as to generate one of the following temperature gradients along the transport path of the transported material flow (2): - in a central portion (4b) of said background (4), the thermal radiation field of said background (4) has a temperature (25) which corresponds at least approximately to the object temperature (26), and in an upstream portion (4a) has a lower temperature than the temperature of said central portion (4b) of said background (4), and in a downstream portion (4c) has a higher temperature than the temperature of said central portion (4b) of said background (4), or - in the central part (4b) of said background (4), the thermal radiation field of said background (4) has a temperature (25) which corresponds at least approximately to the object temperature (26), and in the upstream part (4a) has a temperature which is higher than the temperature of said central part (4b) of said background (4), and in the downstream part (4c) has a temperature which is lower than the temperature of said central part (4b) of said background (4).

10. 10. The method according to claim 1, wherein the temperature control device (6) increases or decreases the background temperature (25) in terms of time or position by an average temperature for that time or position, and further wherein the average temperature for that time or position is changed or adjusted so that at that average temperature the fluctuation amplitude (32) of the measurement signal (29) approaches zero.

11. a tubular product guide is used as background (4), the product guide is rotated about its longitudinal axis and placed in a contrast measurement position, and a contrast measurement is carried out against said background (4) by an infrared sensor (1); The method according to any one of claims 1 to 10, wherein the contrast measurement is carried out while the flow of the conveyed material is continuous or stopped.

12. 12. The method according to claim 1, wherein the measuring spot (20) of the infrared sensor (1) is synchronized with the flow of conveyed material (2) by a rotating prism mirror (24) and / or is guided along the direction of the conveying path of the flow of conveyed material relative to the background (4).

13. the rotation speed of the prism mirror (24) is readjusted based on image evaluation to minimize the average size of the objects (3) in the thermal image of the infrared sensor (1) and / or to produce minimal afterglow tails towards the front and rear, thereby providing single object temperature measurements and, if desired, a statistical description of the temperature uniformity of the product stream; 13. The method according to any one of claims 1 to 12, wherein the measurement spot (20) is moved at an at least approximately constant scanning speed, preferably with a highly dynamically varying angular velocity of the prism mirror (24).

14. 14. The method according to claim 1, wherein the infrared sensor (1) is used to determine the local object temperature and the location of fluctuation minima along the conveying path of the material flow (2) at a plurality of measurement points (51), and the temperature change and path-related temperature change rate of the object (3) are measured from a plurality of measurements along the direction of the material flow (2).

15. 15. The method according to any one of claims 1 to 14, wherein the temperature field of the background (4) varies with position but also with time, and both the positions of variation minima along the direction of the flow of the material (2) conveyed for the object (3) and the object temperatures (39) measured at these positions are recorded, and from the resulting point cloud a path-related temperature change rate of the object (3) is determined, for example by linear regression.

16. 16. The method according to claim 1, wherein, for measuring the temperature of the strands (7) individually, the infrared image is segmented using image processing methods, and separate evaluation sections (51) are defined for individual strands (7) or groups of adjacent strands (7), in particular transversely to the direction of flow of the conveyed material, and the object temperature (39) is determined for each section and / or the temperature gradient in the longitudinal direction of the strand is also determined, additionally with the time variation of the background field (4).

17. 17. The method according to claim 1, wherein a plurality of measurement spots (20), distributed in the direction of flow of the material being conveyed, are captured simultaneously or successively by the infrared sensor (1) by a plurality of sensor elements or an array or matrix of sensor elements.

18. 18. The method according to any of the preceding claims, wherein the position and / or time conditions in which the object (3) has no contrast with the background (4) are used to determine the temperature compensation (40) of the infrared sensor (1).

19. 19. The method according to any of the preceding claims, wherein the infrared sensor (1) operates within a predetermined temperature range that is kept constant under varying ambient conditions and independent of the ambient conditions by an active temperature control device (5), preferably a water temperature control sleeve around the sensor head.

20. In a calibration step, the infrared sensor (1) is relocated from its intended measurement position on the background (4) to a calibration station with an at least substantially black emitter, in which the temperature in the measurement spot is adjusted to an operating point temperature associated with the object to be measured (3), The repositioning is preferably performed without interrupting the power supply and temperature control; The calibration step comprises several sub-steps, namely: First, an infrared detector is aimed at the black emitter of the calibration station and calibrated to a relevant operating point by a first parallel measurement of the infrared signal and the temperature of the calibration reference, and following the absolute operating point calibration, a movement calibration is performed on said background (4), said background (4) being set at least approximately stably to the same operating temperature as the black emitter reference; 20. The method according to claim 1, wherein a second parallel measurement is then performed with the infrared sensor (1) repositioned to the measurement position, wherein the infrared radiation of the background (4) is measured simultaneously and the background temperature (25) is measured by a contact temperature sensor (23), and the degree of radiation of the temperature-controlled background environment as a near-black emitter is determined from the results of the second parallel measurement and saved as a calibration parameter for this operating point.

21. 1. A device for non-contact temperature measurement of filamentous and / or granular objects (3) in a flow of conveyed material (2), comprising: an infrared sensor (1) for capturing the radiation field of a flow of conveyed material (2) flowing against a background (4); a temperature control device (6) that controls the temperature of the background (4) and can generate measurement conditions through the environment, particularly in an environment including specular reflection light; an evaluation device (30) for evaluating the measurement signal (29) from the infrared sensor (1) and for determining the object temperature from the measurement signal (29), the temperature control device (6) is configured to vary the temperature of the background (4) with respect to time and / or position; In the measurement signal (29) from the infrared sensor (1) captured at varying background temperatures (25), the evaluation device (30) evaluates the degree of fluctuation of the signal fluctuation and determines the time and / or position of the fluctuation minimum, It is arranged to determine a determined object temperature (39) from the value of the measurement signal (29) at the time / positional position of the fluctuation minimum. An apparatus characterized in that

22. 22. The apparatus according to claim 21, wherein the evaluation device (30) comprises a regression analysis module (33) suitable for establishing, by means of regression analysis, a functional relationship between the degree of fluctuation (32), in particular the fluctuation amplitude, and time or position, from which the time or position at which a fluctuation minimum, in particular an amplitude minimum, occurs can be determined.

23. 23. The device according to claim 22, wherein in the evaluation device (30) boundaries of temporal and / or locational evaluation zones (45) are defined, which zones have presumably very similar background temperatures (25), preferably characterized by background temperatures (25) within a temperature range or time range or image area, and the evaluation device (30) is preferably configured to calculate and at least temporarily store for each of these evaluation zones (45) the following data for subsequent analysis: - The variability (32) of the measurement signal (29), preferably quantified by evaluating the amplitude, range, trimmed range and / or standard deviation, optionally transformed. - representative temperature reference values ​​selected or calculated by averaging or filtering for an evaluation zone (45), in particular for the background temperature (25), the average measurement signal (35) or the reference temperature (38), which values ​​are as close as possible to the values ​​of the background temperature (25) and the infrared measurement signal (29) in the time and / or location zones where the degree of fluctuation (32) is minimal, and which in other time and / or location zones form a curve as linear as possible to the background temperature (25) with a linear gradient proportional to the difference between the background temperature (25) and the object temperature (26). - a value selected from time to time to represent the evaluation zone (45), preferably characterized by the average value of the start and end times of the data recording of this evaluation zone (45), or a value calculated by averaging or filtering. - A selected value over time, or a value calculated by appropriate averaging or filtering, or a number of values ​​retained for later accurate local analysis, and / or contours (isotherms) delimiting said evaluation zone (45).

24. 24. The apparatus according to any one of claims 21 to 23, wherein the temperature control device (6) is configured to generate a temperature gradient along the conveying path of the conveyed material flow (2): - in a central portion (4b) of said background (4), the thermal radiation field of said background (4) has a temperature (25) which corresponds at least approximately to the object temperature (26), and in an upstream portion (4a) has a temperature which is higher than the temperature of said central portion (4b) of said background (4), and in a downstream portion (4c) has a temperature which is lower than the temperature of said central portion (4b) of said background (4). In the central part (4b) of said background (4), the thermal radiation field of said background (4) has a temperature (25) which corresponds at least approximately to the object temperature (26), in the upstream part (4a) it has a lower temperature than the temperature of said central part (4b) of said background (4) and in the downstream part (4c) it has a higher temperature than the temperature of said central part (4b) of said background (4).

25. Apparatus according to any of claims 21 to 24, wherein a tubular product guide is provided as background (4) and can be rotated about its longitudinal axis and moved into the contrast measurement position.

26. 26. The device according to claim 21, wherein the environment of the measurement zone and the area, which can be indirectly captured by the infrared sensor (1), in particular via the specular reflection angle of the surface of the object (3), emits slightly more infrared radiation than the infrared radiation of the object (3) having the determined object temperature (39) via the housing temperature control (22), thereby compensating for infrared radiation losses in the edge area and the measurement opening (19), generating measuring conditions of specular inclusions in the measurement zone with good approximation, and minimizing the influence of the infrared emissivity ε of the object (3).

27. 27. The device according to any one of claims 21 to 26, wherein the background (4) and / or the inner walls of the housing temperature control (22) are provided with a coating having an emissivity for infrared radiation of more than 40%, more than 70%, or more than 85%.

28. 28. The device according to claim 27, wherein the background (4) and / or the inner walls of the housing temperature control (22) are provided with a non-stick coating of fluoropolymer or silicone.

29. 29. The device according to claim 21, wherein the infrared sensor (1) comprises a plurality of measurement spots (20) along the conveying path of the material flow (2) and / or is capable of determining the local object temperature at a plurality of evaluation points (51).

30. 30. The device according to any one of claims 21 to 29, wherein the infrared sensor (1) comprises a rotating prism mirror (24) for synchronizing the measurement spot (20) with the flow of material being conveyed (2) and / or for guiding the measurement spot (20) along the direction of the conveying path of the flow of material being conveyed relative to the background (4).

31. 31. The device according to claim 30, wherein the control device (30) for controlling the rotation speed of the prism mirror (24) based on the image evaluation is designed in such a way that the average size of the object (3) shown in the thermal image of the infrared sensor (1) is minimized and / or that afterglow tails towards the front and rear are minimized and / or the measurement spot (20) is moved with a highly dynamically changing angular velocity of the prism mirror (24) at an at least approximately constant scanning speed.

32. 32. The device according to claim 21, wherein the infrared sensor (1) comprises a plurality of sensor elements, or a row or matrix of sensor elements, having a plurality of measurement spots (20), arranged in a distributed manner in the direction of flow of the material being conveyed.

33. 33. Apparatus according to any one of claims 21 to 32, wherein an active temperature control device (5), preferably a water temperature control sleeve around the sensor head, is provided for controlling the temperature of the infrared sensor (1) under varying ambient conditions and within a predetermined temperature range that remains constant regardless of the ambient conditions.

34. A granulating device for granulating (pelletizing) plastics, pharmaceuticals, or food products, A granulating apparatus comprising a non-contact temperature measuring device configured according to any one of claims 21 to 33.

35. An underwater pelletizer (13) and a pellet dryer (17) arranged downstream of the underwater pelletizer, 35. A granulating device according to claim 34, wherein the infrared sensor (1) of the non-contact temperature measuring device is directed towards the flow (2) of conveyed material in the dryer outlet (18) of the pellet dryer (17).

36. 35. Granulation device according to claim 34, comprising a strand granulation head for producing strands, wherein an infrared sensor (1) of the device is directed towards the strands (7) emerging from the strand granulation head for contactless temperature measurement.

37. The granulation apparatus of claim 36, wherein the temperature control device (6) is configured to control the temperature of the background (4) behind the filamentous object (3) guided by or flowing around the water only within a narrow horizontal band so that the sensor-weighted infrared radiation corresponds to the required background temperature (25).

Images