阅读说明：本技术 用于检测进给至操作机器的纺织品或金属线的特征的方法、系统和传感器 (Method, system and sensor for detecting characteristics of textile or metal threads fed to a handling machine ) 是由 蒂齐亚诺·巴雷亚 米凯莱·诺尔贾 于 2018-04-11 设计创作，主要内容包括：用于检测进给至操作机器(M)的纺织品或金属线(F)的特征的方法,所述方法包括产生撞击该线的光信号,以便在连接至装置(8)的光学传感器装置上产生阴影,该装置(8)用于基于由这种传感器装置根据通过线(F)本身在所述传感器装置上产生的阴影而发出的电信号来监测线的特征,所述特征是线的物理特征比如线的直径,或是处于运动状态时的线的进给特征比如进给速度。该方法包括：由所述传感器装置(2)所检测的信号包括以模拟模式检测的信号和以数字模式检测的信号,以数字模式检测的信号提供对以模拟模式检测的信号的实时校准,从而产生供监测装置(8)所使用的监测线的特征的电信号。还要求保护根据该方法的操作系统。(Method for detecting the characteristics of a textile or metal thread (F) fed to an operating machine (M), comprising the generation of an optical signal that strikes the thread in order to produce a shadow on an optical sensor device connected to a device (8), the device (8) being intended to monitor the characteristics of the thread, based on an electrical signal emitted by such sensor device as a function of the shadow produced on said sensor device by the thread (F) itself, said characteristics being physical characteristics of the thread, such as the diameter of the thread, or feeding characteristics of the thread when in motion, such as the feed speed. The method comprises the following steps: the signals detected by the sensor means (2) comprise signals detected in an analogue mode and signals detected in a digital mode, the signals detected in the digital mode providing real-time calibration of the signals detected in the analogue mode, thereby generating electrical signals for the characteristics of the monitoring line used by the monitoring means (8). An operating system according to the method is also claimed.)

1. A method for detecting and monitoring a feature of a wire (F) fed to an operating machine (M), the method comprising the optical detection of such a monitored feature, the method providing a first detection of the feature using means of the analog type and a second detection of the feature using means of the digital type for calibrating the detection of the analog type to precisely identify the value of the feature, the first detection of the feature of the analog type being carried out separately from the second detection of the feature of the digital type and generating corresponding data to be sent to a monitoring unit (8), characterized in that a first phase of the monitoring unit (8) determines the value of the feature on the basis of the data detected using means of the digital type, a second phase of the unit varying the data measured by means of the analog type according to the value of the feature determined in the first phase, the modification comprises calibrating the detection of the analog type and determining a determined value of the monitored feature.

2. Method according to claim 1, characterized in that the monitoring unit (8) compares the precise value of the monitored feature with a predetermined value or monitoring parameter of the monitored feature to check whether the precise value corresponds to the predetermined set value; in the case where the precise value does not correspond to the predetermined set value, the monitoring unit (8) generates an alarm signal and/or acts on a feed device that feeds the thread to the machine, so as to prevent the processing of threads having characteristics different from the predetermined set characteristics.

3. Method according to claim 1, characterized in that the modification of the data for the characteristics detected using an analog type device is calculated by a function (C (t)) that varies over time, the value of said function (C (t)) being sampled by the monitoring unit (8) at a frequency higher than 100Hz to 150Hz, advantageously higher than 200Hz, and preferably higher than 300 Hz.

4. The method of claim 1, wherein the calibration is performed in real time.

5. The method according to claim 1, characterized in that the method provides for: the first and second detections are carried out on each of two spatial axes (X, Y) at right angles to each other, the line (F) moving only parallel to one of the axes, the detections on the two spatial axes being able to be compared or crossed with each other to determine a physical characteristic of the line (F).

6. Method according to claim 1, characterized in that the characteristic of the thread detected and monitored is optionally a physical characteristic of the thread (F) or a characteristic associated with the feeding of such thread (F) to a textile machine.

7. The method of claim 6, wherein the physical characteristic is at least one of: -the diameter (F) of the thread, -the fineness, -the local surface variations present in the thread (F) itself, -the knots present along the thread, -the hairiness of the thread (F), -the geometry of the thread (F) or-the number of twists to which the thread is subjected.

8. Method as in claim 6, characterized in that the characteristic associated with the feed of the thread (F) to the textile machine (M) is the feed rate of the thread (F).

9. Method according to claim 5, characterized in that a sequence of at least two illumination phases of the line (F) is specified, one of which comprises zero illumination or darkness.

10. method according to claim 5 and claim 9, characterized in that a sequence of three illumination phases of said line (F) is defined, so as to carry out said first detection and said second detection alternately on one of said axes (X, Y), one of said three illumination phases comprising darkness.

11. A system for detecting and monitoring characteristics of a thread (F) fed to an operating machine (M), said system being able to implement the method of claim 1, said system comprising a detection device (2) of the optical type for the characteristics to be monitored and cooperating with a lighting device (12), said optical detection device (2) comprising detection means (3) operating in analog mode and detection means (4) operating in digital mode, each of said means (3, 4) generating its own electrical signal corresponding to the data of the characteristics to be monitored, the data of the characteristics detected by the detection means (3) operating in analog mode being calibrated by the data of the characteristics detected by the detection means (4) operating in digital mode, a monitoring unit (8) being provided, said monitoring unit (8) being functionally connected to the detection means (3) operating in analog mode and to the detection portion operating in digital mode -a member (4), the monitoring unit (8) being able to receive data about the monitored feature detected by the detection means (3) operating in analog mode and the detection means (4) operating in digital mode, the monitoring unit (8) determining the value of the monitored feature from the data received from the detection means (3, 4), characterized in that the data about the detected feature from the detection means (3) operating in analog mode comprise a portion that varies over time and are unambiguous by the value of the monitored feature determined by the monitoring unit (8) on the basis of the data detected by the detection means operating in digital mode, the determination of the value of the monitored feature enabling calibration of the data of the feature detected by the detection means (3) operating in analog mode, and ultimately the monitored characteristics.

12. System according to claim 10, characterized in that said detection means (3) operating in analog mode and said detection means (4) operating in digital mode belong to a single sensor (2) or to separate sensors (2).

13. System according to claim 11, characterized in that the monitoring unit (8) is connected to a storage unit (9) containing predetermined values of the monitored characteristic, the monitoring unit (8) comparing the determined value of the monitored characteristic with the predetermined values in order to generate a warning and/or to stop the feeding of the wire to the machine when there is a difference between the determined value of the monitored characteristic and the predetermined values.

14. The monitoring system of claim 11, wherein the illumination device is a single light source.

15. System according to claim 11, characterized in that the detection means (3) operating in analog mode and the detection means (4) operating in digital mode, the monitoring unit (8), the storage unit (9) and the illumination means (12) for the detection means (3, 4) are parts of a single body (5) which are hit simultaneously by the shadow cast by the thread (F) when the thread is illuminated by the illumination means (12).

16. System according to claim 11, characterized in that a pair of detection members operating in analog mode (3) and in digital mode (4) is provided, positioned on each of two axes (X, Y) spatially at right angles to each other, each of said pair of detection members cooperating with a respective lighting device (12), the wire (F) fed to the operating machine being moved between the detection member operating in analog mode (3) and the detection member operating in digital mode (4).

17. System according to claim 16, characterized in that on one (X) of said spatial axes (X, Y) there are two detection means (3A, 3B) in analog mode and one detection means operating in digital mode (4), said detection means being at a short distance from each other.

18. The system according to claim 11, characterized in that said monitored characteristic is optionally a physical characteristic of said wire (F): such as the diameter, fineness, surface deformation, geometry, twist count, interlacing, hairiness or knotting of the thread (F), or a characteristic associated with the feeding of the thread (F), i.e. the feed rate.

Technical Field

The present invention relates to a method for detecting characteristics of a textile or wire fed to a textile machine according to the preamble of the main claim. A system for detecting the characteristics of such a line and operating according to said method and a sensor for use in such a system according to the respective independent claims also comprise the object of the present invention.

Background

As is known, at least one characteristic of the conventional routing, which may be dimensional or related to the feed (such as the feed speed), is monitored by suitable sensor means when feeding the thread into the operating machine. This makes it possible to leave the operating machine with a final product having characteristics and quality that meet predetermined specifications.

Herein, the term "thread" is intended to mean both a woven thread or yarn as well as a metal thread; similarly, the term "operating machine" covers both textile machines (looms, beamers, knitting machines, cross-winders or other machines) and machines that perform any operation on the wire, such as winding the wire on a spool.

In this context, the term "end product" is intended to cover both textile goods and products containing the thread being monitored, whether textile or metal (such as bobbins).

The term "characteristics of the thread" refers to any dimensional characteristics of the thread (such as diameter or fineness or local thinness), local surface deformations (filaments or hairiness (hairs) protruding from the surface of the fiber), changes in the linearity of the fiber (such as winding itself into a loop), or even characteristics associated with the motion of the thread, such as the feed speed of the thread to the operating machine.

Various methods and devices are known for detecting the above-mentioned characteristics of the thread as it is fed to the operating machine. This document relates to devices and apparatus of the above type that operate by using an optical system to determine the characteristics.

Devices are known which are capable of checking the characteristics of the fed thread, provided for the said use of the device of optical sensors. Such sensors work on the principle of physical measurement, which consists in estimating the size of the shadow produced by the wire on the detector component of such a sensor, after it has been "hit" (struck) by the light emitted by the lighting means, such as an LED. Based on this estimate, it is possible to determine, for example, the diameter of the thread or whether there are any surface irregularities or knots in the thread.

In general, there are currently two types of sensors that are mutually alternative, namely sensors operating in analog mode or digital sensors (that is, in general, these sensors operate by detection of the analog or digital type, respectively).

Analog sensors typically include a photodiode on which a shadow of a line struck by light generated by an LED is formed. Assuming that the illumination is planar and uniform, the ratio of the thickness or width "d" of the shadow of the line to the width "W" of the photodiode is proportional to the amount of light measured by the sensor. Fig. 1 shows a typical photodiode S on which a shadow O of a line F is formed; dimensions d and W are shown in the figure, as well as the length or height H of the photodiode.

Based on these values, the diameter d of the wire can be obtained from a simple ratio

(H-d)/H＝P_MIS/P₀

Where P0 is the power measured without wires, and P_MISIs the measured power with the line. This is converted into

d＝H(1-P_MIS/P₀)

However, analog sensors are not able to make highly accurate determinations, and the measurements of analog sensors depend on many factors. For example, in the case of illumination of a line, this illumination must be well collimated (neither divergent nor convergent) so that the size of the shadow of the line produced on the photodiode does not vary with the distance between the line and the photodiode.

It follows that for example the individual light sources have to be located far apart, which is however not possible in most applications where the detector arrangement comprising photodiodes and LEDs has to be of very small size and thus the distance between light source and detector is small. This problem can be overcome by: a dedicated (cylindrical) lens placed between the LED and the photodiode is used in order to produce at least one collimated illumination beam in a direction at right angles to the line, but an illumination beam diverging in a parallel direction may also be produced.

It is also possible to provide a compensation algorithm applied to the data detected by the photodiodes used by the units monitoring the above data (connected to the photodiodes), but this results in the need to process such data and therefore makes the monitoring unit more complex, and thus eventually the means of detecting the characteristics of all the lines comprise the LEDs, the photodiodes and said monitoring unit.

This method of estimating the diameter is simple and fast, but presents various accuracy problems due to the possibility of variations in the light source over time, spatial non-uniformities, ambient light, the possibility of the line being transparent, the non-uniform response of the photodiode, and errors associated with the hairiness (hairiness) of the line or the unevenness of its surface.

Devices that operate using digital sensors and overcome many of the problems of devices using analog sensors are also known.

It is well known that digital methods can overcome many of these problems. If the analog sensor is replaced with a sensor array (typically manufactured using CMOS or CCD technology), the position of the resulting shadow edge determined algorithmically can be used to more accurately estimate the diameter of the wire. The geometry of the sensor itself, with a resolution of the order of a few microns (pixel size), guarantees the accuracy of the measurement.

In this respect, various solutions are known for measuring the characteristics of the wire by means of a sensor array, as described above.

GB-2064106 describes a CCD device comprising around one hundred optical sensors, said device being designed to analyse a filiform image and being able to determine the diameter of a wire. The shadow produced by the line is scanned at points along the length of the CCD, that is, the exposure state of each of the optical sensors is checked in turn. Thus, the CCD generates a pulse sequence for each scan period. The diameter of the wire is thus continuously converted into a large number of serial pulses.

US-4511253 also describes a linear array of optical sensors and a circuit for evaluating the serial signal provided by the linear array.

WO-9936746 also describes a CCD sensor and a method for determining the thickness of a wire.

Unlike the previous patent documents which refer to measuring devices using CCD sensors, WO-2011147385 describes the preferred use of nmos (live mos), JFET, LBCSAST and Scmos sensors, since the low consumption of these sensors is an advantage when they are incorporated into small devices, and they do not require a cooling system.

EP-1319926 describes a device intended to measure at least one characteristic of a wire, such as for example the diameter. Also, the optical sensor used here is a CMOS sensor.

EP-2827127 describes an optical sensor comprising two parallel rows of optical elements. The optical elements in the first row are rectangular and oriented such that the long sides of the optical elements are along the direction of the projected movement of the line. The optical elements in the second row also have a rectangular shape, but are oriented such that the long sides of these optical elements are perpendicular to the direction of the projected movement of the line. In this case, all optical elements are also constructed using CMOS technology.

JP-S60114704 does not precisely relate to a wire but to a cable or the like, and describes a method of measuring the diameter of a cable or the like based on the width of a shadow detected by an optical sensor, the width of the shadow being kept constant even if the distance between the cable and the optical sensor varies.

For this particular application, the main drawback of such sensors is the measurement speed. In contrast to analog sensors, the number of signals that must be acquired from the sensor array is equal to the number of its pixels. The larger the number of pixels, the slower the measurement of the line characteristics, compared to the case where the line characteristics are obtained using a single analog sensor. This can be overcome by connecting the digital sensors to (final) devices for measuring the monitored characteristics, which have a high calculation speed (operating on the basis of the signals emitted by these sensors), but this would make the assembly or system thus obtained more costly. Therefore, the application of digital sensors generating signals, for example also corresponding to measurements of the diameter of wires in tens of thousands per second, cannot be associated with the application of low-cost digital electronics or microcontrollers, or else it would be unlikely that a cost-acceptable detection system would be produced.

GB 2159621 relates to a method and apparatus for monitoring the size of a plurality of products, both stationary and in motion. This prior patent document describes that the product that has to be scanned is irradiated by a laser beam and the transmitted light produced is collected by a photocell. At the same time, the product is also illuminated by another beam of light at another location, and the light passing through the product is collected by another photocell. The first photocell generates a pulse signal, which is denoted as a digital signal in this patent document, while the second photocell generates an analog signal; these signals are processed separately by (different) electronic circuits, thereby generating two versions of a measurement of the mobile product.

The two signals are added and the result is an analog value for accurate diameter measurement.

Thus, this prior document describes the detection of analog data and digital data, the analog data being calibrated by the digital data. However, this detection is achieved by using two separate light sources at two different points on the product, which makes the known device described in GB 2159621 very complex and increases the size of the known device.

In addition to this, in the prior art document, ordinary illumination is used to obtain an analogue measurement of the diameter of the product and this may cause problems in the measurement itself, since such illumination may vary, which may result in erroneous measured diameter values.

WO 00/62013 relates to a method and apparatus for measuring the diameter of a transparent fibre and monitoring the surface defects of the transparent fibre. This prior art technique involves measuring the above-mentioned characteristics of the optical fibre by detecting the number of interference fringes rather than by measuring the geometrical shadow produced by the product being measured on the detector. Since the measurement is based on interferometry and the diameter of the fiber (transparent fiber) is measured, the known invention requires the use of coherent light that must be generated by a laser. Therefore, this known solution is very complex (since it also detects internal defects in the fibre or "air duct") and costly both because of the mode of operation (based on interferometry) and also because of the method used to obtain such measurements and monitoring of the product. This solution therefore requires a device for analyzing the data obtained, which is very complex and costly.

The known solutions are therefore not usable to detect the diameter of the thread in textile machines or operating machines that treat textiles or metal threads, due to their complexity (and therefore to their size) and to their cost.

US 6219135 describes a device for optically detecting at least one parameter, such as the diameter, of a moving filamentary material. The device comprises an optical sensor having two detector components or a single sensor, a first single sensor operating in an analog mode and a second single sensor operating in a digital mode. The thread or filamentary material moves between the optical sensor and the light source, which may also be a direct light source or a light source capable of producing reflected light before illuminating the thread.

The data detected by each single sensor is processed by an evaluation circuit which generates a signal proportional to the diameter of the wire or filamentary material. The accuracy of the measurement depends on the number of individual sensors used per unit length of the optical sensor, or on whether the individual signals generated by the individual sensors are modulated (processed in analog mode) or recorded only in binary form (in this way obtaining digital signals).

Also, the prior document describes that a value of the hairiness of the thread can be obtained by comparing an analog signal with a digital signal.

This prior document is silent about whether and how the data detected in analog mode and the data detected in digital mode interact when determining the characteristics (e.g. diameter) of the measured filamentary material.

The prior literature is likewise silent about the scanning times of the products.

EP 2423144 describes a device for detecting the movement of a wire, which comprises detecting the aforementioned diameter using sensors positioned non-uniformly at a distance from each other. Comparing the similarity of the values determined by the two sensors in order to clarify the similarity between the detected data; these similarities are then weighted against each other.

Provision is made for a delay between the signals with weighted similarity detected by the above-mentioned weighted sensors, and information relating to the movement of the line is determined on the basis thereof.

This solution describes the use of an optical sensor to determine the speed of the line movement.

CH 671041 describes an electro-optical sensor device for the geometry of wires, which uses an electro-optical sensor to generate an analog signal, which is digitized after filtering.

A unit for detecting moir é defects is provided.

This prior document describes converting an analog signal to a digital signal rather than calibrating the analog signal with a digital signal.

Disclosure of Invention

the object of the present invention is to provide a method and a system implementing the method and a sensor for the system, by means of which the characteristics of a textile or wire fed to an operating machine can be determined quickly, accurately and at low cost.

In particular, it is an object of the present invention to provide a method, system and sensor of the above-mentioned type, by means of which one or more dimensional characteristics of the wire, such as the diameter, fineness, hairiness, geometry and twist number of the wire, can be detected.

Another object is to provide a method, system or sensor of the above type which is capable of detecting the feed rate of the thread.

Another object is to provide a system of the above type which is small in size so as to facilitate its application in textile machines operating hundreds of threads simultaneously.

Another object is to provide a method, system and sensor of the above-mentioned type, by which ambient light can be used directly to measure the characteristics of a line without the need to use infrared radiation.

These and other objects that will be apparent to a person skilled in the art are achieved by a method, a system and a sensor according to the respective independent claims.

Drawings

For a better understanding of the invention, the following drawings are attached thereto by way of non-limiting example only; in the drawings:

FIG. 1 is a schematic diagram showing a known analog sensor in detecting dimensional characteristics of a line F;

FIG. 2 is a schematic diagram illustrating a system according to the present invention;

FIG. 3 is a schematic diagram showing a first variation of the system of FIG. 2; and

Fig. 4 is a schematic diagram showing a second variation of the system in fig. 2.

Detailed Description

With reference to the figures, and in particular to fig. 2, there is shown a system 1 for detecting characteristics of a thread F fed to an operating machine M. Such a line functions with an optical sensor 2 which, in the example in the figures, has an analogue detection member 3 and a digital detection member 4, the analogue detection member 3 and the digital detection member 4 being associated with a single container body 5, and the analogue detection member 3 and the digital detection member 4 being positioned at a short distance from each other in this body 5. This makes it possible to have a detection system 1 (or detection unit or detection device) of very small dimensions, so that this detection system 1 can be applied, together with other identical systems, to each thread to be monitored in textile machines operating on hundreds of threads, such as knitting machines and the like.

When said detection means 3 and 4 are also present in the body 5, compactness of the system 1 can also be achieved due to the fact that: the fact that only 1 light source or LED (not shown in fig. 2) is able to "hit" the line F and allow the shadow to be produced on both of the detection means 3 and 4, allowing the means to emit an electrical signal corresponding to the size of such shadow. In particular, the analog detection component 3 may be a photodiode, while the component 4 may be a CMOS or CCD sensor, or similar semiconductor sensor. The component 4 uses a vector sensor to spatially digitize the shadow of the resulting line.

The semiconductor sensors in the component 4 define a matrix of photodetectors by means of which the above-mentioned spatial digitization can be achieved.

As will be described below, thanks to the use of these semiconductor sensors, it is possible to obtain "instantaneous photographs" of the wire in a few microseconds, regardless of the feed speed of the pipe.

It is also possible to determine the diameter of such a thread accurately, when it moves towards the textile machine (or winding machine for metal threads or similar operating machine).

For example, accurate measurements cannot be made using the description described in WO 00/62013, because according to this prior document, as already described, the characteristics of the wire are measured by interferometry; since the wire is in motion and therefore vibrates, determining the interference fringes in a manner that is useful for measuring the diameter of the wire may be at least as of say imprecise.

As occurs in WO 00/62013, the diameter of the wire can be determined immediately by the resulting shadow of the wire on the component 4 without the need to process the signal.

The detection means 3 and 4 (or analog sensor 3 and digital sensor 4, respectively) are connected to a monitoring or evaluation unit 8 (also present in the body 5) of the microprocessor type, the monitoring or evaluation unit 8 calculating the characteristics (for example, the diameter) of the monitored thread F on the basis of the electrical signals or data emitted by these means 3 and 4. Unit 8 receives and analyzes signals originating from both sensor 3 and sensor 4.

The units 8 are connected to a storage unit 9, in which storage unit 9 predetermined acceptance values (or monitoring parameters) for monitoring features are inserted, and these units 8 are used to compare the found data or actual data with the predetermined acceptance values in order to evaluate the correspondence of said data with predetermined values; if the actual value does not correspond to the value in the memory, the unit 8 acts in a known manner, for example to prevent the machine M from continuing to use the thread F with characteristics different from those desired, by generating a visible and/or audible warning, generating a signal to the device feeding the thread F to the machine M or other known devices.

The system 1 provides measurements (in the form of electrical signals) by the digital part 4 of the sensor 2 and provides real-time calibration of the measurements (also in the form of electrical signals) in the analog part 3. It should be noted that the term "calibration" means periodically determining the measurement from the analog part 3 of the sensor 2 and comparing it with the value of the diameter measured by the digital part 4 of the sensor.

The unit 8 uses the "calibrated" electrical signals originating from the components 3 and 4 to carry out the above estimation.

Also taking into account possible errors in the analog sensor, the diameter measurement made by such a sensor or analog component 3 can be expressed in short term by a linear equation in terms of the power measured by the analog component 3.

d＝H(1-C(t)·P_MIS)

Wherein, P_MISis the value of the power measured by the analog part 3 (photodiode), which varies with time, H is the known dimension of the analog part 3, and C is a variable as a function of time (we define it as the "calibration variable") and depends on a number of factors: the optical power of the illumination, the presence or absence of dirt, the transparency of the wire, and any change in the gain of the electronics (taking into account that the background value will vary with temperature). The value of the variable c (t) over time is estimated at a predetermined frequency higher than 100Hz to 150Hz, advantageously higher than 200Hz, preferably 300 Hz. By making such a periodic estimation at the above-mentioned sampling frequency pair c (t), factors that may adversely affect the value of c (t), i.e. factors that vary at a very low frequency, as described above, can be overcome. Vice versa, the above-mentioned frequencies for timely estimation of the c (t) value in the precise instance can be used to determine the dimension value of the line, since at these frequencies the only component that can vary rapidly is the lateral position of the line itself. It has in fact been found that in practical cases the wire may move linearly or even laterally in case of vibrations. However, since it has been found that the vibration frequency is at most equal to a few tens of hertz, these movements do not affect the measurement if the sampling is performed at a frequency higher than 100Hz to 50 Hz; as mentioned above, such vibrations thus have no effect on the value of c (t) when measured.

In other words, it has been found that the magnitude of the vibrations affecting the wire in its movement from the bobbin (unwinding from the bobbin) to the operating machine (textile machine or machine operating on the wire) is at most 10 Hz. The detected signal is sampled using a frequency of at least 100Hz and preferably higher than 300Hz, the image of the line detected by the analog part 3 of the sensor 2 necessarily displaying the line as if it were completely stationary, in order to allow the characteristics of the line, in particular the diameter of the line, to be determined.

It should be noted that, as will be pointed out, the signals detected by the components 4 are also sampled at the same frequency, which also makes it possible to detect the characteristics of the lines, as is still the case with digital components or sensors 4.

At the right time of unit 8The value of the diameter d1 of the line is determined on the basis of the data detected by the digital part 4 of the sensor 2 (associated with the shadow produced on this part 4) at a particular time (and at a continuous and discrete timing frequency as described above). Once this measurement is completed, a formula similar to the one described above can be applied and thanks to the dimensions H of the analog sensor 3 and the power P measured by it_MISBoth are known, so the change in calibration C can be calculated as

C(t)＝(1-d1/H)/P_MIS

When the value of c (t) at each instant in the measurement time is obtained, the value may be inserted into a formula for calculating "d" by using an analog sensor, thereby calibrating the signal detected by the analog sensor.

In summary, if the calibration parameter or variable C is calculated by means of the diameter d1 measured by the digital part 4 of the sensor 2, the above problematic effects (based on the above reasons) can be compensated for in the case of, for example, at least 100Hz (or higher). Such a measurement rate can be easily achieved using CMOS sensors and low cost electronics.

The method according to the invention therefore provides for: the analog part 3 of the sensor 2 independently detects, in a manner known per se, the characteristics (for example the diameter) of the monitored wire and generates its own detection signal and sends it to the unit 8. The measurement of the component 3 using the sensor 2 is fast, but as is known, the measurement is inaccurate.

In parallel, the digital part 4 of the sensor 2 detects the above-mentioned characteristic (d1), also independently and in a manner known per se, and generates its own signal and sends it to the unit 8. The latter uses the above-described method to calibrate the signal coming from the component 3 by means of the signal generated by the digital component 4, which is more accurate than the signal coming from the analog component and is generated at a low sampling speed, so that a very small (commercially acceptable) and very low-cost microprocessor unit 8 (or "microcontroller") can be used.

Based on the "calibrated" signal, the unit 8 has the role of comparing this signal with the data stored in the memory and, if there is any difference from the latter, the unit 8 generates an alarm or acts in the manner described above.

Fig. 3 shows a system for measuring the fineness of a thread F requiring two measuring axes (X and Y in fig. 3), wherein parts corresponding to those in the already described figures are denoted by the same reference numerals. In this case, it is proposed to use two sensors 2, the light rays of the two sensors 2 generated by the LEDs 10 being at right angles to one another. Also shown in the figure is a semi-cylindrical lens 12 which directs light to the sensor 2 and the cell 8, the cell 8 being considered to include the above-mentioned circuitry 20. The LED is also connected to the unit 8.

These sensors 2, LEDs and units 8 are all associated with a single body 5.

A system for measuring fineness requires two measurement axes, which can be achieved by repeating the method in two dimensions. To avoid interference from ambient light, the system may be pulsed and selectively detected by both photodiodes and CMOS sensors. To minimize measurement time, a sequence of three illumination states may be provided: the first axis is bright, and the second axis is bright and dark. In this way, the measured difference with respect to the darkness makes it possible to eliminate the interference of the ambient light without having to resort to optical filters, thus making it possible to work also in visible light.

One example of using the system in fig. 3 is as follows. It is expected that measurements at a frequency of 33kHz, which can produce a single light pulse of 10 mus duration, will provide a total time to measure background brightness of 30 mus from two pulses (one pulse per axis) and a dark of 10 mus. Two 512 pixel CMOS sensors are sampled at 1MHz, obtaining a sampling time for each CMOS of approximately 0.5ms, and 1ms for both sensors if collected in sequence. 0.5ms is sufficient for parallel acquisition. The time for digitally processing the data to calculate the measured diameter must be increased to this time and is limited to about 2ms for a low cost microcontroller.

In summary, measurements made by two photodiodes at about 300Hz can be calibrated, and this includes a 33kHz data stream that is sufficient to detect knots or defects.

In summary, in both cases described (fig. 2 and 3), the analog sensor guarantees the measurement rate, while the digital sensor guarantees the measurement accuracy.

Fig. 4 shows a system 1, in fig. 4 parts corresponding to those of the already described figures are denoted by the same reference numerals, which system 1 can also be used to determine the feed speed along the line F of the X-axis (or along an axis at right angles to the Y-axis).

With regard to such embodiments, it is known to monitor anomalies related to physical characteristics of the wire (e.g., the diameter or fineness of the wire) typically based on predetermined monitoring parameters or values such as percentage and length.

For example, by increasing the fineness by 50% on a 1mm long line, the presence or absence of knots can be detected.

This means that the monitoring unit (e.g. unit 8) inspecting the characteristics of the line must operate based on a suitable algorithm, which is also based on knowledge of the feed rate of the line, so that it is possible to calculate in real time how long it takes for an anomaly in the line to be detected in order to have a monitored line length of 1 mm.

In the system according to the invention, it is therefore also of utmost importance to be able to determine the speed of the thread as it is fed into the textile machine reliably, in real time and as accurately as possible.

One of the advantages of being able to calculate the linear speed in real time is that, in this way, the speed is no longer included in the set of parameters that must be programmed to monitor the feed of the thread to the textile machine. Furthermore, the system also successfully protects the monitoring threshold from the machine speed, ensuring the measurement of the length required to evaluate the fineness of the wire and/or the anomalies in the diameter of the wire, even during the acceleration phase or deceleration phase or during the operator's change of the operating speed.

The speed signal may also be used as a synchronization signal to enable or disable monitoring, for example enabling monitoring at speeds above 300 m/min.

In this way, knowledge of the speed makes the system completely independent, but does not require any signal to synchronize with the machine and any synchronization signal from the machine.

According to the variant of the invention in question, the speed is determined by comparing the signals generated by a pair of analog components 3A and 3B present in the sensor 2 located on and operating along the X axis. Knowing the distance between the analog components and determining the delay for the time of detection of a particular characteristic of the wire (e.g., a change in hairiness, diameter, or other characteristic), the feed rate of the wire can be determined using known mathematical formulas relating to time, distance, and speed. The most reliable known technique for this determination consists in calculating a correlation function between the two analog signals, which function will show a peak corresponding to the detection time delay in the case of a line with minimum surface defects.

In any case, only the data from the analog of the two components 3A and 3B of the sensor 2 is used to detect the characteristics of the monitored line (according to the description with respect to fig. 2 and 3).

Various embodiments of the present invention have been described. However, other embodiments, such as providing for the use of two different sensors (always associated with a single supporting body), each with its own LED and its own detector component, are also possible, the embodiment in which the first sensor (photodiode) operates in analog mode and the other sensors (CMOS or CCD) operate in digital mode. However, also in this case, the signal from the "digital sensor" is used to calibrate the signal emitted by the "analogue sensor" (with reference to the nature of the detector components) before the monitoring unit 8 determines whether the monitored characteristic value is acceptable (within predetermined parameters).

These variants are able to provide sensors that can be used in systems operating according to the method described to solve the above technical problem, and also have the characteristics described in the following claims.