Roller rotating speed monitoring device and monitoring method based on reflective mark

Abstract

The invention relates to the technical field of roller rotating speed monitoring, in particular to a roller rotating speed monitoring device and monitoring method based on a reflective mark, and the device comprises the reflective mark, an intelligent light source regulation and control system and a light spot visual acquisition and processing unit. The reflective mark is of a three-layer composite structure, the intelligent light source regulation and control system can dynamically adjust the modulation frequency, and the light spot visual acquisition and processing unit acquires and processes light spot images through a high-speed industrial camera and an image processor; according to the method, light source intelligent modulation and synchronization are carried out by arranging reflective marks, the real-time rotating speed is calculated by using a multi-dimensional information fusion algorithm, and an alarm mechanism is set; the method is high in anti-interference capability, high in measurement precision, capable of adapting to different working conditions, low in cost and easy to implement, and the problems of inaccurate monitoring, high cost, poor adaptability and the like in the prior art are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of roll speed monitoring, in particular to a roll speed monitoring device and method based on reflective markers. BACKGROUND

[0002] In the rough rolling production process, the operator needs to observe the running state of the roll in real time, including whether it rotates and the rotating speed, etc. At present, the commonly used method is to make marks on the surface of the roll, and to judge the running state of the roll by observing the movement of the marks. However, in the actual production environment, there are often contaminants such as oil stains, water stains and oxide scales on the surface of the roll, which can cover or blur the marks, making it difficult for the operator to clearly observe the movement of the marks, so as to accurately judge the rotating speed of the roll, and even may cause production accidents due to misjudgment.

[0003] In addition, the existing method relies on subjective judgment of human vision and lacks quantitative standards. When the roll speed is in a critical state, it is easy to misjudge due to the visual persistence effect. The physical durability of the traditional marks is insufficient, and under the impact of high-speed rolling and frequent rolling load, the marks are prone to wear, fall off and other problems, which requires frequent downtime maintenance and affects production efficiency.

[0004] Some schemes try to use infrared sensors or laser Doppler velocimeters for non-contact measurement, but infrared sensors are easily disturbed by environmental temperature, and the measurement accuracy is greatly reduced under high-temperature working conditions of the roll. The laser Doppler system is high in cost, and the installation position and light path alignment are required to be very high, which is difficult to popularize in complex rolling mill sites. The above-mentioned technologies need to modify the electrical control system of the rolling mill, which is difficult to implement and has poor compatibility.

[0005] At the same time, some working conditions do not allow the use of speed measurement equipment with encoders: 1) space limited, unable to install encoders; 2) high-temperature environment beyond the tolerance range of encoders; 3) high precision requirements, not allowed to have any additional devices fall off, otherwise it may cause equipment damage or major production accidents. Therefore, there is an urgent need for a more reliable, more intuitive, lower-cost and non-mechanical contact speed measurement method to solve the problem of rough rolling roll speed monitoring. SUMMARY

[0006] In view of the deficiencies of the prior art, the present application aims to provide a roll speed monitoring device and method based on reflective markers, which is low in installation cost, intuitive and reliable.

[0007] The technical scheme adopted by the present application to solve its technical problems is:

[0008] A roller speed monitoring device based on reflective marking includes a reflective marking, an intelligent light source control system and a light spot visual acquisition and processing unit; wherein the reflective marking is pasted on the roller, and the reflective marking is a multi-layer composite structure, including three layers, from the inside to the outside, a high-temperature resistant silicone rubber adhesive layer, a microprism reflective structure layer and a surface super-oleophobic nano-coating; the intelligent light source control system includes multiple groups of frequency-modulated LED light source modules, each group of frequency-modulated LED light source modules includes a main light source, an auxiliary light source and a light source controller, and the light source controller is connected to the main light source and the auxiliary light source; the light spot visual acquisition and processing unit includes a high-speed industrial camera and an image processor, the image processor is connected to the high-speed industrial camera and the light source controller, and the high-speed industrial camera is used to collect the light spot image reflected by the reflective marking on the roller surface; the data processed by the image processor is output to an external monitoring system or a host computer through a data interface.

[0009] Preferably, a further technical solution of the present invention is:

[0010] Preferably, the reflective mark includes a plurality of reflective units, and the plurality of reflective units are arranged in a spiral structure to form an array of reflective units distributed at equal angles in the circumference of the roller.

[0011] Preferably, the arc length L of a single reflective unit satisfies the formula:

[0012]

[0013] Where D is the roller diameter, z is the number of reflective units, and k is the spiral coefficient, which ranges from 0.8 to 1.2.

[0014] Preferably, the number of reflective units is calculated using the following formula:

[0015]

[0016] Among them, k2 is the unit selection coefficient, which ranges from 100 to 200;

[0017] The axial spacing d between adjacent reflective units satisfies:

[0018]

[0019] Among them, v max is the maximum measurable speed, f max is the maximum frame rate of the camera, k3 is the redundancy coefficient, and the value range is 3-6.

[0020] Preferably, the light source modulation frequency f when the device is first operated is init and the roll design speed n des Satisfies the formula:

[0021]

[0022] Wherein, z is the number of light reflection units, m is the sampling rate, and the value range is 2-5.

[0023] Preferably, the main light source is a red light LED with a wavelength of 650nm, the auxiliary light source is an infrared light supplement lamp with a wavelength of 850nm, and the light source controller is provided with a PWM modulation circuit, which can realize pulse frequency adjustment of 10-1000Hz.

[0024] The application further discloses a rolling mill roller rotating speed monitoring method based on the light reflection mark.

[0025] Arranging the light reflection mark: the light reflection mark is pasted on the surface of the rolling mill roller;

[0026] Intelligent modulation and synchronization of light source:

[0027] In the starting stage, f init Modulate the light source, and the high-speed industrial camera collects the light spot image to estimate the initial rotating speed estimation value n0;

[0028]

[0029] Wherein, v time0 is the initial linear velocity calculated by the light spot displacement in the starting stage; then the reference frequency f0 is calculated:

[0030]

[0031] Wherein, k1 is a correction coefficient, and the value is 0.8-1.2;

[0032] In the steady state stage: the light source is based on the reference frequency f0, and the modulation frequency is dynamically adjusted according to the real-time rotating speed n real ;

[0033]

[0034] Wherein, n base is the preset rotating speed of the process; and n real is solved by a multi-dimensional information fusion algorithm of time domain, frequency domain and brightness characteristics.

[0035] Specifically, n real The calculation method is,

[0036] Time domain feature solving: the light spot displacement Δp of adjacent frames is calculated, and the real-time linear velocity v time is converted through the calibration coefficient K:

[0037] v time = Δp * K (8) ;

[0038] Wherein, the calculation formula of the calibration coefficient K is:

[0039]

[0040] where N pix is the number of pixels corresponding to the roll circumference; v time is the real-time linear velocity of the roll; real_time

[0041]

[0042] Frequency domain feature solution: perform FFT transformation on the spot brightness sequence, extract the main frequency f peak , and calculate the frequency domain linear velocity v freq ;

[0043]

[0044] Frequency domain component n real_freq of the real-time rotational speed is inversely calculated:

[0045]

[0046] Brightness feature solution: LSTM network based on deep learning is used to fit the linear velocity v light :

[0047] v light = f(I, T, θ) (13);

[0048] where I is the average spot brightness value, T is the ambient temperature obtained through the temperature sensor arranged on the spot, and θ is the observation angle;

[0049] Brightness component n real_light of the real-time rotational speed is inversely calculated:

[0050]

[0051] The time domain feature velocity v time , the frequency domain feature velocity v freq , and the brightness feature velocity v light are weighted and fused to obtain the final real-time rotational speed n real of the roll;

[0052] n real = ω1×n real_time + ω2×n real_freq + ω3×n real_light (15);

[0053] where ω1, ω2, and ω3 are the time domain feature weight, the frequency domain feature weight, and the brightness feature weight, respectively.

[0054] ​Preferably, ω1, ω2, ω3 are dynamically adjusted according to preset working condition parameters, the working condition parameters including a facula brightness fluctuation coefficient γ, a noise standard deviation σ and an observation angle offset θ, wherein:

[0055] The facula brightness fluctuation coefficient γ is a ratio of an absolute value of a difference between a current frame facula brightness and a reference brightness to the reference brightness;

[0056] When γ≤0.3 and σ≤0.5, ω1 takes a value of 0.4-0.5, ω2 takes a value of 0.3-0.4, and ω3 takes a value of 0.1-0.2;

[0057] When γ>0.3 or σ>0.5, ω1 takes a value of 0.5-0.7, ω2 takes a value of 0.3-0.5, and ω3 takes a value of 0.05-0.1.

[0058] The application also sets an alarm threshold λ, when the real-time rotating speed n real is greater than λ or the facula disappears for more than 5 frames, the monitoring system or the upper computer alarms.

[0059] The application adopting the above technical scheme has the following prominent features compared with the prior art:

[0060] 1. Strong anti-interference capability: the reflective identification adopts a three-layer composite structure, the surface layer of super oil-repellent nano coating can resist pollutants such as oil stains and oxide scales, the micro-prism reflective structure layer ensures the reflective effect, the high-temperature resistant silicone rubber bonding layer guarantees the stability of the paste under high-temperature working conditions, and the light source configuration combining red light and infrared light improves the monitoring reliability in complex environments;

[0061] 2. High measurement accuracy: the real-time rotating speed is calculated through a multi-dimensional information fusion algorithm of time domain, frequency domain and brightness characteristics, and the wavelet transform is used for denoising, and the dynamic weighted adjustment is combined to effectively improve the speed measurement accuracy and meet the high-precision monitoring demand under different working conditions.

[0062] 3. Good adaptability: the intelligent light source control system can dynamically adjust the modulation frequency according to the real-time rotating speed to adapt to the change of the rolling mill rotating speed; the spiral arrangement of the reflective unit and the related parameter calculation formula ensure the monitoring effect in different diameters of the rolling mill and the rotating speed range.

[0063] 4. Low cost and easy implementation: no modification is needed for the electrical control system of the rolling mill, and the cost is low compared with the laser Doppler scheme, and the installation and maintenance are convenient. BRIEF DESCRIPTION OF DRAWINGS

[0064] Figure 1 is a structural schematic diagram of a rolling mill rotating speed monitoring device in the embodiment of the application;

[0065] Figure 2 is a schematic diagram of the arrangement of the reflective unit in the embodiment of the application.

[0066] Fig. 1, roll; 2, light reflection unit. DETAILED DESCRIPTION

[0067] The present application is further described below in conjunction with specific embodiments, which are intended to better illustrate the present application and are not intended to limit the scope of protection of the present application.

[0068] As Figure 1 The present embodiment provides a roll rotational speed monitoring device based on light reflection identification, which comprises a light reflection identification, an intelligent light source control system, and a light spot visual acquisition and processing unit. The light reflection identification is pasted on the roll 1. The light reflection identification is a multi-layer composite structure, which comprises three layers, i.e., a high-temperature-resistant silicone rubber bonding layer, a micro-prism light reflection structure layer, and a surface super-oleophobic nano coating layer from inside to outside. The high-temperature-resistant silicone rubber bonding layer can withstand a high temperature of 200-300° and has a bonding force with the surface of the roll 1 of ≥5 N / cm 2 ; the light reflection efficiency of the micro-prism light reflection structure layer is ≥80% under 650 nm red light.

[0069] The intelligent light source control system comprises multiple groups of tunable LED light source modules. Each group of tunable LED light source modules comprises a main light source, an auxiliary light source, and a light source controller. The light source controller is connected to the main light source and the auxiliary light source. The main light source is a red light LED with a wavelength of 650 nm, and the auxiliary light source is an infrared supplementary light with a wavelength of 850 nm. The light source controller is internally provided with a PWM modulation circuit, which can realize pulse frequency adjustment of 10-1000 Hz. The light spot visual acquisition and processing unit comprises a high-speed industrial camera and an image processor. The image processor is connected to the high-speed industrial camera and the light source controller. The high-speed industrial camera is used to acquire the light spot image reflected by the light reflection identification on the surface of the roll 1. The data processed by the image processor is output to an external monitoring system or an upper computer through a data interface. The frame rate of the high-speed industrial camera is not less than 1000 fps, and the resolution is not less than 1280×720. The image processor is integrated with an FPGA chip.

[0070] Specifically, the light reflection identification comprises multiple light reflection units 2, which are arranged in a spiral structure and form an equiangularly distributed light reflection unit 2 array in the circumferential direction of the roll 1. The arc length L of a single light reflection unit 2 satisfies the formula:

[0071]

[0072] wherein D is the diameter of the roll 1, z is the number of light reflection units 2, and k is the spiral coefficient, which is in the range of 0.8-1.2.

[0073] The number of specifically arranged light reflection units 2 is calculated using the following formula:

[0074]

[0075] Wherein, k2 is a unit selection coefficient, taking 100-200;

[0076] The axial spacing d of adjacent light reflection units 2 satisfies:

[0077]

[0078] Wherein, v max is the maximum measurable speed, f max is the maximum frame rate of the camera, and k3 is a redundancy coefficient, taking 3-6.

[0079] The light reflection unit 2 arrangement example is as follows: taking the roll 1 diameter of 600 mm as an example, the design speed range n des = 100-800 r / min, the unit selection coefficient k2 is 150, and z is equal to 25; taking the spiral coefficient k = 1, the arc length L of a single light reflection unit 2 is 75.4 mm. The industrial camera is selected to be Basler acA2500-14gm, f max = 1400 fps, the pixel angular resolution is 0.5°, and the redundancy coefficient k3 = 5; the axial spacing d of the light reflection unit 2 is 28.6 mm.

[0080] The light source modulation frequency f init of the device when it is first operated and the roll 1 design speed n des satisfy the formula:

[0081]

[0082] Wherein, z is the number of light reflection units 2, and m is the sampling rate, taking the value range of 2-5.

[0083] The application further discloses a roll speed monitoring method based on a light reflection mark, which is applied to the roll speed monitoring device based on the light reflection mark, and the specific steps are as follows:

[0084] Arranging the light reflection mark: the light reflection mark is pasted on the surface of the roll 1;

[0085] Intelligent light source modulation and synchronization:

[0086] In the starting stage, the light source is modulated at f init , and the high-speed industrial camera collects the light spot image to estimate the initial speed estimation value n0;

[0087]

[0088] Wherein, v time0 is the initial linear speed calculated by the light spot displacement in the starting stage; then the reference frequency is calculated:

[0089]

[0090] wherein k1 is a correction coefficient, taking a value of 0.8-1.2;

[0091] In the steady state phase: the light source is based on the reference frequency f0, according to the real-time speed n real dynamically adjust the modulation frequency;

[0092]

[0093] wherein n base is the process preset speed; n real is solved by a multi-dimensional information fusion algorithm of time domain, frequency domain and brightness characteristics.

[0094] Specifically, n real The calculation method is,

[0095] Time domain feature solving: calculate the adjacent frame light spot displacement Δp, and convert it into real-time linear speed v time :

[0096] v time = Δp × K (8) ;

[0097] wherein the calculation formula of the calibration coefficient K is:

[0098]

[0099] wherein N pix is the pixel number corresponding to the circumference of the roll; from the real-time linear speed v time , the real-time speed time domain component n real_time is inversely deduced:

[0100]

[0101] Frequency domain feature solving: perform FFT transformation on the light spot brightness sequence, extract the main frequency f peak , and calculate the frequency domain linear speed v freq ;

[0102]

[0103] The real-time speed frequency domain component n real_freq is inversely deduced:

[0104]

[0105] Brightness feature solving: the linear speed v light is fitted based on the LSTM network of deep learning:

[0106] v light = f (I, T, θ) (13) ;

[0107] Wherein, I is the average spot brightness value, T is the ambient temperature, which is obtained by arranging the temperature sensor on the spot; θ is the observation angle;

[0108] Backstepping real-time speed brightness component n real_light :

[0109]

[0110] The time domain characteristic velocity v time , the frequency domain characteristic velocity v freq and the brightness characteristic velocity v light are weighted and fused to obtain the final rolling mill 1 operation real-time speed n real .

[0111] n real = ω1×n real_time + ω2×n real_freq + ω3×n real_light (15).

[0112] Wherein, ω1, ω2, ω3 are respectively the time domain characteristic weight, the frequency domain characteristic weight and the brightness characteristic weight.

[0113] Further, ω1, ω2, ω3 are dynamically adjusted according to the preset working condition parameters, and the working condition parameters include the spot brightness fluctuation coefficient γ, the noise standard deviation σ and the observation angle offset θ, wherein:

[0114] The spot brightness fluctuation coefficient γ is the ratio of the absolute value of the difference between the current frame spot brightness and the reference brightness to the reference brightness;

[0115] When γ≤0.3 and σ≤0.5, ω1 takes the value of 0.4-0.5, ω2 takes the value of 0.3-0.4, and ω3 takes the value of 0.1-0.2;

[0116] When γ>0.3 or σ>0.5, ω1 takes the value of 0.5-0.7, ω2 takes the value of 0.3-0.5, and ω3 takes the value of 0.05-0.1.

[0117] It is verified by experiments that, under the working condition that the surface temperature of the rolling mill 1 reaches 250℃ and the oil stain coverage reaches 30%, the retroreflective sign can still maintain the reflection efficiency >75%. When the spot brightness fluctuation coefficient γ>0.4, the system automatically increases the time domain weight to 0.7 and reduces the frequency domain weight to 0.3, so as to ensure that the speed calculation error is <±0.5%.

[0118] The application also sets an alarm threshold λ, when the real-time speed n realWhen the light spot is greater than λ or disappears for more than 5 frames, the monitoring system or the upper computer alarms. In an example operation, when a certain reflective unit 2 is manually blocked, if the light spot is not detected for 6 consecutive frames, the upper computer displays an "light spot loss" alarm and automatically increases the light source power by 30% (the main light source power to 2.6W).

[0119] Intelligent early warning mechanism: three-level alarm system (±5% speed fluctuation warning, ±10% speed reduction, 5-frame light spot loss shutdown) combined with automatic fault handling, significantly reduces the risk of production accidents.

[0120] The present application can effectively resist pollutants such as oil stains and oxide scales by using a three-layer composite reflective marker, and can still maintain a reflectivity of ≥75% under high temperature (200-300℃) and polluted environment, ensuring monitoring stability. Through the multi-dimensional information fusion algorithm of the LSTM model combined with dynamic weight adjustment, the speed calculation error is <±0.5%, which is significantly better than the traditional infrared sensor (error ±5%). Through the intelligent light source regulation system, the modulation frequency is dynamically adjusted according to the real-time speed, and the environmental interference is automatically compensated (such as increasing the light source power by 30% when the light spot is lost), which is suitable for sudden changes in the speed of the roller 1 or harsh working conditions.

[0121] The device does not need to modify the electrical system of the rolling mill, uses conventional LED and fill light, and can be realized by combining a high-speed industrial camera, and the comprehensive cost is only 10% of the laser Doppler scheme.

[0122] Experiments show that in the case of a 600mm diameter roller 1 and a speed of 100-800r / min, the present application can still maintain stable monitoring under high temperature and high pollution conditions, solving the problems of traditional methods such as easy blurring of markers, high manual misjudgment rate and poor equipment compatibility.

[0123] The above only describes the preferred embodiments of the present application, and is not intended to limit the scope of the present application, and any equivalent changes made by applying the content of the present application specification and drawings are included in the scope of the present application.

Claims

1. A roll rotational speed monitoring device based on retro-reflective markers, characterized by: The roller speed monitoring device based on the reflective identification comprises a reflective identification, an intelligent light source control system and a light spot visual acquisition and processing unit; the reflective identification is pasted on the roller; the reflective identification is a multi-layer composite structure comprising three layers, from inside to outside, a high-temperature-resistant silicone rubber bonding layer, a micro-prism reflective structure layer and an oil-repellent nano coating layer on the surface; The intelligent light source control system comprises multiple groups of frequency-adjustable LED light source modules; each group of frequency-adjustable LED light source modules comprises a main light source, an auxiliary light source and a light source controller; the light source controller is connected to the main light source and the auxiliary light source; The light spot visual acquisition and processing unit comprises a high-speed industrial camera and an image processor; the image processor is connected to the high-speed industrial camera and the light source controller; the high-speed industrial camera is used to acquire the light spot image reflected by the reflective identification on the surface of the roller; the data processed by the image processor is output to an external monitoring system or an upper computer through a data interface.

2. The retro-reflective marker based roll rotational speed monitoring device according to claim 1, characterized in that: The reflective identification comprises a plurality of reflective units; the plurality of reflective units are arranged in a spiral structure and form an equiangularly distributed reflective unit array in the circumferential direction of the roller.

3. The retro-reflective marker based roll rotational speed monitoring device of claim 2, wherein: The arc length L of a single reflective unit satisfies the formula: wherein D is the diameter of the roller, z is the number of reflective units, and k is a spiral coefficient, which is in the range of 0.8-1.

2.

4. The retro-reflective marker based roll rotational speed monitoring apparatus according to claim 3, characterized by The number of reflective units is calculated using the following formula: wherein k2 is a unit selection coefficient, which is in the range of 100-200; The axial spacing d of adjacent reflective units satisfies: where v max is the maximum measurable speed, f max is the maximum frame rate of the camera, and k3 is a redundancy coefficient, which takes a value between 3 and 6.

5. The retro-reflective marker based roll rotational speed monitoring apparatus according to claim 1, characterized in that: The light source modulation frequency f at first operation of the device init With the roll design rotational speed n des Satisfies the formula: wherein z is the number of reflective units, and m is a sampling rate, which is in the range of 2-5.

6. The retro-reflective marker based roll rotational speed monitoring apparatus according to claim 1, characterized in that: The main light source is a red light LED with a wavelength of 650 nm, and the auxiliary light source is an infrared light supplement lamp with a wavelength of 850 nm; the light source controller is provided with a PWM modulation circuit, which can realize pulse frequency adjustment in the range of 10-1000 Hz.

7. A method for monitoring the rotational speed of a roll based on retro-reflective markers, characterized in that The roller speed monitoring device based on the reflective identification is applied to the method according to any one of claims 1-6, and the specific steps are as follows: arranging the reflective identification: pasting the reflective identification on the surface of the roller; intelligent modulation and synchronization of the light source In the start-up phase, f init Modulate the light source, high-speed industrial camera collects light spot image, estimates the initial rotation speed estimate n0; where v time0 is the initial linear velocity for the start-up phase spot displacement calculation; then the reference frequency is calculated: wherein k1 is a correction coefficient, which is in the range of 0.8-1.2; In the steady state phase: the light source is based on the reference frequency f0, according to the real-time speed n real Dynamic adjustment of the modulation frequency; wherein n base is the preset rotation speed of the process; n real Solved by a multi-dimensional information fusion algorithm of time domain, frequency domain and brightness characteristics.

8. The retro-reflective marker based roll rotational speed monitoring method of claim 7, wherein, n real The calculation method is that, Time domain feature solution: calculate the adjacent frame light spot displacement Δp, and convert it into real-time linear velocity v through calibration coefficient K time : v time = Δp x K (8); wherein the calculation formula of the calibration coefficient K is: N = (L - 2 * R) / 2 pix N is the pixel number corresponding to the roll circumference; the real-time linear speed v time Backstepping real-time speed time domain component n real_time : Frequency domain feature solution: the spot brightness sequence is transformed by FFT, and the main frequency f is extracted peak , and the frequency domain linear velocity v is calculated freq ; Counter-rotating real-time rotational speed frequency domain component n real_freq : Brightness feature solution: LSTM network based on deep learning to fit linear velocity v light : v light = f(I, T, θ) (13); wherein I is the average brightness value of the light spot, T is the environmental temperature, which is obtained by a temperature sensor arranged on the site, and θ is the observation angle; Back-calculation real-time rotational speed brightness component n real_light : The time domain characteristic velocity v time , the frequency domain characteristic velocity v freq and the brightness characteristic velocity v light are weighted and fused to obtain the final rolling mill running real-time rotating speed n real ; n real = ω1 x n real_time + ω2 x n real_freq + ω3 x n real_light (15); wherein ω1, ω2 and ω3 are respectively a time domain feature weight, a frequency domain feature weight and a brightness feature weight.

9. The retro-reflective marker based roll rotational speed monitoring method of claim 8, wherein: ω1, ω2 and ω3 are dynamically adjusted according to preset working condition parameters, which include a light spot brightness fluctuation coefficient γ, a noise standard deviation σ and an observation angle offset θ, wherein: the light spot brightness fluctuation coefficient γ is the ratio of the absolute value of the difference between the current frame light spot brightness and the reference brightness to the reference brightness; when γ≤0.3 and σ≤0.5, ω1 is in the range of 0.4-0.5, ω2 is in the range of 0.3-0.4, and ω3 is in the range of 0.1-0.2; when γ>0.3 or σ>0.5, ω1 is in the range of 0.5-0.7, ω2 is in the range of 0.3-0.5, and ω3 is in the range of 0.05-0.

1.

10. The retro-reflective marker based roll rotational speed monitoring method of claim 8, wherein: An alarm threshold λ is set, when the real-time rotating speed n real When the rotating speed is greater than λ or the light spot disappears for more than 5 frames, the monitoring system or the upper computer alarms.

Images