Iron ore belt sampling method and system

Abstract

This invention designs a method and system for iron ore conveyor belt sampling, belonging to the field of iron ore sampling technology. It acquires real-time cross-sectional contour data of the material flow using a 3D laser contour scanning unit and a surface texture imaging unit. Based on the cross-sectional contour data, it calculates the instantaneous virtual mass flux and the material flow morphology dispersion. The sampling trigger threshold is adaptively adjusted based on the material flow morphology dispersion. The instantaneous virtual mass flux is integrated to obtain the cumulative flux, and sampling is triggered when the threshold is reached. The lateral movement speed of the sampling head is dynamically adjusted according to the material layer thickness distribution to achieve variable-speed cutting. Closed-loop correction is performed through weighing feedback. The system includes a 3D laser contour scanning unit, a surface texture imaging unit, an intelligent sampling control terminal, a frequency converter-driven sampler, and a weighing feedback unit. This invention solves the sampling deviation problem caused by uneven material flow distribution and improves sample representativeness.

Description

Technical Field

[0001] This invention relates to the field of iron ore sampling technology, specifically to an iron ore belt sampling method and system. Background Technology

[0002] Iron ore is a fundamental raw material for the steel industry, and its quality directly affects the performance and production cost of steel products. In port iron ore inspection and quarantine, sampling is a crucial preliminary step for quality assessment and trade settlement, and the accuracy of the sampling results directly impacts the economic interests of both trading parties. Iron ore sampling must comply with relevant national and international standards, with the core objective of obtaining sufficiently representative samples whose chemical composition and particle size distribution accurately reflect the actual condition of the entire shipment.

[0003] Port iron ore sampling is broadly categorized into manual sampling and mechanical sampling based on the operation method. Manual sampling is suitable for small batches of goods or areas inaccessible to machinery, but it is less efficient and easily affected by human error. Mechanical sampling is the mainstream method for large-scale continuous operations, offering advantages such as high efficiency and low error. Belt samplers are commonly used in mechanical sampling. They automatically capture samples during the conveyor belt transport of iron ore. The sampling head crosses the belt to laterally intercept the material flow, collecting the intercepted material into a sampling hopper to form a sample. During belt conveying, iron ore undergoes segregation due to factors such as vibration and tilt. This manifests as large particles tending to roll to the sides of the belt while fine powder concentrates in the middle, and the thickness of the material flow fluctuates along the conveying direction. This non-uniformity in spatiotemporal distribution poses challenges to sampling. Traditional belt sampling methods typically employ timed or quantitative sampling triggering methods, i.e., triggering sampling at fixed time intervals or fixed cumulative tonnage. This fixed-threshold sampling triggering method does not consider the time-varying characteristics of material flow. When the material flow fluctuates drastically, the sampling timing does not match the changes in material flow characteristics. When the material flow is stable, it may generate unnecessary sampling times, resulting in insufficient representativeness of the sample for material flow changes. Traditional belt samplers typically use a constant speed to cross the belt for cutting. Because the material layer is unevenly distributed along the belt width, the amount of material to be cut per unit time increases when the sampling head passes through areas with thicker material layers. The sampling hopper may overflow due to instantaneous filling. The overflowing material is often dominated by large particles, resulting in a low content of large particles in the sample and affecting particle size representativeness. The material layer is thicker and rich in large particles in the off-center loading area, and the uniform speed cutting method is particularly prone to overflow problems in this area. In addition, traditional sampling systems usually lack effective feedback correction mechanisms, and the system parameters are difficult to adaptively adjust according to actual working conditions. After long-term operation, the sampling accuracy may gradually deviate from the set target.

[0004] To address the aforementioned issues, a belt sampling method and system is needed that can sense the cross-sectional morphological characteristics of the material flow in real time and dynamically adjust the sampling strategy accordingly, in order to improve the accuracy of the sampling results and the representativeness of the samples. Summary of the Invention

[0005] To address the shortcomings of existing technologies, the technical solution adopted by this invention to solve its technical problems is as follows:

[0006] A method for sampling iron ore conveyor belts includes the following steps:

[0007] S1. By installing a 3D laser contour scanning unit and a surface texture imaging unit upstream of the sampling point of the belt conveyor, the cross-sectional contour data and surface texture image data of the iron ore flow are acquired in real time.

[0008] S2. Based on the cross-sectional profile data, combined with the real-time running speed of the belt and the preset bulk density parameters, calculate the instantaneous virtual mass flux at the current moment.

[0009] S3. Calculate the material flow pattern dispersion based on the cross-sectional profile data, and adaptively compensate and adjust the preset sampling trigger threshold based on the material flow pattern dispersion to obtain the dynamic sampling trigger threshold.

[0010] S4. Integrate the instantaneous virtual mass flux over time to obtain the cumulative flux. When the cumulative flux reaches the dynamic sampling trigger threshold, send a sampling trigger signal to the sampler.

[0011] S5. Obtain the material layer thickness distribution in the belt width direction based on the cross-sectional profile data, and dynamically adjust the transverse movement speed of the sampling head based on the material layer thickness distribution so that the sampling head can perform transverse cutting at different speeds in different material layer thickness areas.

[0012] S6. The single sampling quantity is weighed and verified through the weighing feedback unit, and the weighing data is fed back to the intelligent sampling control terminal for the correction and optimization of sampling parameters.

[0013] As a preferred embodiment, a further technical solution of the present invention is:

[0014] In a further embodiment, step S1 includes:

[0015] S11. Install the 3D laser contour scanning unit 5 to 10 meters upstream of the sampling point of the belt conveyor, and scan the cross-sectional contour of the iron ore flow on the belt in real time at a preset scanning frequency to obtain the material layer height value at each position in the width direction of the belt, and form the cross-sectional height contour function h(x,t).

[0016] S12. Install a high-speed industrial camera adjacent to the 3D laser contour scanning unit to acquire iron ore surface texture images at a preset frame rate to help determine the particle size segregation situation.

[0017] S13. The optical window of the 3D laser contour scanning unit is continuously purged by an air knife purging device to prevent dust from adhering and affecting the scanning accuracy.

[0018] S14. The acquired cross-sectional height profile function h(x,t) and surface texture image data are transmitted to the intelligent sampling control terminal in real time to complete the acquisition of material flow cross-sectional morphology data.

[0019] In a further scheme, in step S2, the formula for calculating the instantaneous virtual mass flux Φ(t) is:

[0020] ;

[0021] In the formula, Φ(t) is the instantaneous virtual mass flux at time t (kg / s); ρ is the preset bulk density of iron ore (kg / m³). 3 v(t) is the real-time running speed of the belt at time t (m / s); h(x,t) is the material layer height at position x in the width direction of the belt at time t (m); W is the effective width of the belt (m). The integral of the cross-sectional profile height function along the belt width direction represents the cross-sectional area (m²). 2 ).

[0022] In a further embodiment, step S3 includes:

[0023] S31. Based on the cross-sectional profile data at multiple consecutive sampling times, calculate the ratio of the standard deviation to the mean of the cross-sectional area sequence to obtain the material flow pattern dispersion ΔS.

[0024] S32. Based on the material flow pattern dispersion ΔS, set the preset sampling trigger threshold Q. set Adaptive compensation adjustment is performed, and the trigger threshold Q is dynamically sampled. dynamic The calculation formula is:

[0025] Q dynamic =Q set ×(1+α×ΔS);

[0026] In the formula, Q dynamic The dynamic sampling trigger threshold (kg); Q set The preset sampling trigger threshold (kg) is α, which is the dispersion compensation coefficient, ranging from 0.1 to 0.5; ΔS is the material flow pattern dispersion, a dimensionless parameter.

[0027] S33. When the dispersion of material flow pattern ΔS increases, the dynamic sampling trigger threshold increases accordingly to capture the fluctuation characteristics of material flow; when the dispersion of material flow pattern ΔS decreases, the dynamic sampling trigger threshold approaches the preset sampling trigger threshold, thus completing the adaptive adjustment of the sampling trigger threshold based on the material flow pattern characteristics.

[0028] In a further embodiment, step S4 includes:

[0029] S41, the time t when the previous sampling was completed last Starting with the instantaneous virtual mass flux Φ(t), we perform continuous-time integration to obtain the cumulative flux M(t). The formula for calculating the cumulative flux M(t) is as follows:

[0030] ;

[0031] In the formula, M(t) is the cumulative flux (kg) at the current time t; t last τ is the time (s) when the last sampling was completed; Φ(τ) is the instantaneous virtual mass flux (kg / s) of the integral variable at time τ.

[0032] S42. Real-time comparison of cumulative flux M(t) and dynamic sampling trigger threshold Q. dynamic Size;

[0033] S43, When M(t) ≥ Q dynamic At that time, the intelligent sampling control terminal sends a sampling trigger signal to the frequency converter driven sampler, and simultaneously resets the cumulative flux M(t) to zero and updates t. last At the current moment, a dynamic sampling trigger based on virtual quality integral is completed.

[0034] In a further embodiment, step S5 includes:

[0035] S51. Divide the belt width direction into N equally spaced intervals, and calculate the average material layer thickness h(y) of each interval and the average material layer thickness H of the entire cross section based on the cross-sectional profile data. avg ;

[0036] S52. Calculate the transverse velocity v of the sampling head in each interval based on the material layer thickness h(y). cutter (y), lateral velocity v cutter The formula for calculating (y) is:

[0037] ;

[0038] In the formula, v cutter (y) represents the lateral velocity (m / s) of the sampling head at position y of the belt width; v base The reference lateral velocity (m / s); H avg ε is the average material layer thickness across the entire cross section (m); h(y) is the material layer thickness at position y (m); ε is the zero constant for damping, ranging from 0.001 to 0.01 (m); β is the damping coefficient, ranging from 0.3 to 0.5.

[0039] S53. During the process of the sampling head crossing the belt, the intelligent sampling control terminal adjusts the lateral movement speed v. cutter (y) Real-time control of the motor speed of the variable frequency drive sampler, so that the sampling head can complete variable speed cutting and sampling based on the cross-sectional thickness distribution.

[0040] In a further embodiment, step S5 also includes:

[0041] S54. The calculated lateral velocity v cutter (y) performs boundary constraints, when v cutter (y) is greater than the maximum allowable lateral speed of the motor, v. max At that time, v cutter (y) is restricted to v max When v cutter (y) is less than the minimum lateral velocity v min At that time, v cutter (y) is restricted to v min ;

[0042] S55. When a single cut is insufficient to completely capture the material in a localized section of a material layer, the system automatically performs two cuts to complete the sampling of the localized section of the material layer, ensuring the integrity and representativeness of the sampling.

[0043] In a further embodiment, step S6 includes:

[0044] S61. Install a weighing sensor inside the sampling hopper or an electronic belt scale on the sampling conveying path to weigh the sample obtained each time in real time, and obtain the measured sampling quantity M. real ;

[0045] S62. The measured sample quantity M real Compare the sample size with the theoretical sample size and calculate the sampling deviation rate;

[0046] S63. Feedback the sampling deviation rate to the intelligent sampling control terminal. When the sampling deviation rate exceeds the preset threshold, automatically adjust the preset bulk density parameter ρ or the dispersion compensation coefficient α to complete the closed-loop correction and optimization of the sampling parameters.

[0047] This invention also discloses an iron ore conveyor belt sampling system, comprising:

[0048] The 3D laser contour scanning unit is installed upstream of the sampling point of the belt conveyor to acquire cross-sectional contour data of the iron ore flow in real time.

[0049] The surface texture imaging unit is installed adjacent to the 3D laser contour scanning unit to acquire surface texture image data of iron ore.

[0050] The intelligent sampling control terminal is connected to the 3D laser contour scanning unit and the surface texture imaging unit respectively. It receives cross-sectional contour data and surface texture image data, calculates instantaneous virtual mass flux, cumulative flux and material flow pattern dispersion, and adaptively adjusts the sampling trigger threshold based on the material flow pattern dispersion, as well as calculates the dynamic lateral movement speed of the sampling head according to the cross-sectional thickness distribution.

[0051] The variable frequency drive sampler is connected to the intelligent sampling control terminal, receives sampling trigger signals and dynamic transverse speed commands, and executes variable speed cutting sampling actions.

[0052] The weighing feedback unit is connected to the intelligent sampling control terminal to verify the weighing quantity and feed back the weighing data.

[0053] In the preferred embodiment, the 3D laser contour scanning unit is installed 5-10 meters upstream of the sampling point on the belt conveyor to acquire cross-sectional morphology data of the material flow before sampling. The 3D laser contour scanning unit is equipped with an air knife blowing device to continuously blow the optical window to prevent dust adhesion. The surface texture imaging unit is a high-speed industrial camera to assist in judging the particle size segregation of iron ore and provide image basis for calculating the transverse segregation index. The intelligent sampling control terminal includes an image processing module and a sampling decision module.

[0054] The image processing module processes cross-sectional contour data and surface texture image data to extract material flow morphology feature parameters; the sampling decision module calculates the dynamic sampling trigger threshold and dynamic lateral movement speed based on the material flow morphology feature parameters; the variable frequency drive sampler includes a sampling head and a variable frequency drive motor; the opening size of the sampling head is not less than three times the maximum particle size of the iron ore; the variable frequency drive motor is connected to the intelligent sampling control terminal for control, and adjusts the lateral movement speed of the sampling head according to the dynamic lateral movement speed command to achieve variable speed cutting during the sampling head's traversal of the belt; the weighing feedback unit includes a weighing sensor installed in the sampling hopper or an electronic belt scale installed on the sampling conveyor path, which acquires the single sampling volume in real time and feeds the weighing data back to the intelligent sampling control terminal for closed-loop correction; it also includes a belt speed detection unit, which is connected to the intelligent sampling control terminal signal to acquire the belt running speed in real time, providing speed parameters for the calculation of instantaneous virtual mass flux.

[0055] The present invention, which adopts the above technical solution, has the following prominent features compared with the prior art:

[0056] First, this invention employs an indirect calculation method based on geometric measurement. By measuring the cross-sectional area of ​​the material flow and combining it with the material bulk density and belt speed, it estimates the instantaneous virtual mass flux. This fully utilizes the cross-sectional morphology data already acquired by the 3D laser contour scanning unit, eliminating the need for additional dynamic metering equipment. This avoids the problem of limited measurement accuracy of traditional dynamic metering equipment under vibration environments. The air knife blowing device continuously blows the optical window of the 3D laser contour scanning unit, effectively blocking and removing high-concentration dust particles in the port environment, keeping the optical channel clean, ensuring that the measurement accuracy is not affected by dust, and improving the reliability and stability of the system under harsh working conditions.

[0057] Secondly, this invention introduces a material flow morphology dispersion parameter. Based on cross-sectional contour data from multiple consecutive sampling times, the ratio of the standard deviation to the mean of the cross-sectional area sequence is calculated to characterize the fluctuation of the material flow within the observation time window. The preset sampling trigger threshold is adaptively compensated and adjusted based on the material flow morphology dispersion to obtain a dynamic sampling trigger threshold. When the material flow fluctuates drastically, the dynamic sampling trigger threshold increases accordingly, and the system delays sampling to wait for more material to pass through, thereby capturing the fluctuation characteristics of the material flow over a longer time window and improving the representativeness of the sample to material flow changes. When the material flow is stable, the dynamic sampling trigger threshold approaches the preset threshold, and the system samples at a normal pace, avoiding unnecessary sampling. This adaptive adjustment mechanism matches the sampling frequency with the material flow characteristics, optimizing sampling efficiency while ensuring sample representativeness. It solves the problem of mismatch between sample collection timing and material flow characteristic changes caused by the traditional fixed threshold sampling method's failure to consider the time-varying characteristics of the material flow morphology. This invention employs a sampling trigger control method based on cumulative flux. The instantaneous virtual mass flux is integrated over time to obtain the cumulative flux. Sampling is triggered when the cumulative flux reaches the dynamic sampling trigger threshold. This ensures that the material batch represented by each sampling is basically consistent, avoiding the problem of uneven sampling intervals caused by fluctuations in material flow rate. It meets the national standard requirement that the sampling quantity should be proportional to the material batch, effectively improving the stability of the sampling quantity.

[0058] Third, the iron ore conveyor belt sampling method provided by this invention obtains the material layer thickness distribution along the width direction of the conveyor belt based on cross-sectional profile data. Based on this distribution, the lateral movement speed of the sampling head is dynamically adjusted. This allows the sampling head to cut at a lower speed in areas with thicker material layers to fully absorb the material, and at a higher speed in areas with thinner material layers to improve efficiency. During conveyor belt transport, due to the combined effects of the material's inherent characteristics and conveyor vibration, large particles tend to roll to the sides of the belt, creating an off-center load. In these off-center load areas, the material layer is thicker and larger particles are concentrated. Traditional uniform-speed cutting methods in these areas result in excessively fast cutting speeds, causing the sampling hopper to be unable to absorb the entire layer of material, leading to overflow. The missed material is mainly large particles, resulting in a low content of large particles in the sample and affecting particle size representativeness. This invention effectively solves the problem of missed large particles caused by off-center loading through a variable-speed cutting sampling method, improving the particle size representativeness of the sample. This invention also incorporates a speed boundary limit and a fault-tolerant protection mechanism for staged cutting. When the calculated lateral speed exceeds the motor's allowable speed range, boundary limits are automatically applied. When the thickness of a localized area of ​​the material layer is too large to completely capture the entire layer in a single cut, the system automatically performs two cuts to complete the sampling of that area, ensuring the integrity and representativeness of the sampling while protecting the safe operation of the equipment. This invention uses a weighing feedback unit to verify the weight of each sample and feeds the weighing data back to the intelligent sampling control terminal for parameter correction and optimization, forming a closed-loop control system. This ensures that the system parameters continuously approximate actual operating conditions, improving the accuracy and stability of the system's long-term operation.

[0059] Fourth, the iron ore conveyor belt sampling system provided by this invention includes a 3D laser contour scanning unit, a surface texture imaging unit, an intelligent sampling control terminal, a frequency converter-driven sampler, a weighing feedback unit, and a conveyor belt speed detection unit. These units are connected via signal cables or industrial communication networks, forming a complete control closed loop encompassing data acquisition, analysis and processing, control execution, and feedback correction. The 3D laser contour scanning unit employs a combination of a line laser and a high-speed image sensor. It calculates the height coordinates of each point on the material flow surface using triangulation principles, accurately acquiring the height contour distribution of the material flow cross-section. The surface texture imaging unit uses a high-speed industrial camera in conjunction with a ring-shaped LED illumination device to acquire clear images of the material flow surface, assisting in determining the iron ore particle size segregation and providing image data for calculating the transverse segregation index. The intelligent sampling control terminal includes an image processing module and a sampling decision module. The image processing module processes the cross-sectional contour data and surface texture image data and extracts material flow morphology characteristic parameters. The sampling decision module calculates the dynamic sampling trigger threshold and dynamic transverse speed based on the material flow morphology characteristic parameters and generates corresponding control commands, achieving integrated processing of material flow morphology data analysis and processing with sampling control decision generation. The variable frequency drive sampler includes a sampling head and a variable frequency drive motor. The opening size of the sampling head is not less than three times the maximum particle size of the iron ore, which complies with the national standard for sampler opening size. This ensures that materials of various particle sizes can enter the sampling hopper without obstruction, avoiding sampling deviation caused by large particles clogging the sampling port. The variable frequency drive motor adjusts the motor speed according to the dynamic lateral movement speed command, realizing variable speed cutting during the sampling head's traversal of the belt. Compared with traditional uniform speed samplers, it has higher speed control accuracy and faster response speed.

[0060] Fifth, this invention solves the problem of sampling deviation caused by uneven material flow distribution in traditional belt conveyor sampling technology by combining a dynamic sampling triggering strategy based on virtual mass integral and a variable-speed cutting sampling method based on cross-sectional thickness distribution. The dynamic sampling triggering strategy ensures that the sampling timing matches the material flow state, and the sampling stability under fluctuating material flow conditions is significantly better than that of timed sampling methods, meeting the standard requirements for the coefficient of variation of mechanical sampling quality. The variable-speed cutting sampling method ensures the representativeness of sample particle size, and under off-center loading conditions, the particle size representativeness is significantly better than that of uniform-speed cutting methods, effectively avoiding particle size deviation caused by overflow. This ensures that the sampling deviation rate meets the international standard requirements for sampling precision in iron ore trade, improving the overall representativeness of the sample for the material batch. The system of this invention achieves intelligent and adaptive control of the iron ore belt conveyor sampling process through a complete control loop of perception, decision-making, execution, and feedback. The closed-loop control architecture formed by the collaborative work of each unit can automatically adjust sampling parameters and sampling actions according to real-time changes in material flow morphology, adapting to the sampling needs of different working conditions and different iron ore varieties, and has good versatility and adaptability. Attached Figure Description

[0061] Figure 1 This is a flowchart of an iron ore belt sampling method according to an embodiment of the present invention. Detailed Implementation

[0062] The present invention will be further illustrated below with reference to specific embodiments. The purpose of this illustration is solely to provide a better understanding of the invention. Therefore, the examples given do not limit the scope of protection of the present invention.

[0063] During iron ore conveying, factors such as vibration and tilting cause material segregation, where large particles tend to roll to the sides of the belt, while fine powder concentrates in the middle, and the material thickness fluctuates along the conveying direction. Traditional timed or quantitative sampling methods do not consider this spatiotemporal non-uniformity, easily leading to systematic sampling bias and affecting sample representativeness. (Reference) Figure 1 This invention provides a method for iron ore belt sampling. Addressing the problem of uneven material flow distribution leading to sampling deviations in traditional belt sampling techniques, this method utilizes machine vision technology to perceive the cross-sectional morphology of the material flow in real time. Based on this morphology, the method dynamically adjusts the sampling trigger timing and the lateral movement speed of the sampling head, achieving adaptive control of the sampling process. By constructing a virtual quality model of the material flow cross-section, the method couples the material flow morphology with the sampling control strategy, improving the accuracy of the sampling results.

[0064] The iron ore conveyor belt sampling method of the present invention includes the following steps: Step S1 is the acquisition of material flow cross-sectional morphology data. A 3D laser contour scanning unit and a surface texture imaging unit installed upstream of the sampling point on the conveyor belt acquire cross-sectional contour data and surface texture image data of the iron ore flow in real time. The 3D laser contour scanning unit uses the laser triangulation principle or time-of-flight ranging principle to emit a laser beam onto the conveyor belt surface. By receiving the reflected light signal, it calculates the spatial coordinates of the intersection point between the laser beam and the material surface, thereby obtaining the height contour distribution of the material flow cross-section. The surface texture imaging unit uses a high-speed industrial camera to continuously photograph the material flow surface, acquiring texture feature images of the material surface. These images can be used to assist in determining the particle size distribution and segregation state of the material. Installing the sensors upstream of the sampling point allows the system to acquire material flow morphology information before the material reaches the sampling point, providing data preparation time for subsequent sampling decisions.

[0065] Step S1 specifically includes the following sub-steps. Step S11 involves installing the 3D laser contour scanning unit at a preset distance upstream of the sampling point on the belt conveyor. This preset distance is preferably 5 to 10 meters and can be adjusted according to the belt speed and system data processing time. The purpose is to ensure that the system completes the acquisition and processing of material flow morphology data before the material reaches the sampling point. The 3D laser contour scanning unit scans the cross-sectional contour of the iron ore flow on the belt in real time at a preset scanning frequency, obtaining the material layer height value at each position in the belt width direction, forming a cross-sectional height contour function h(x,t), where x represents the position coordinate in the belt width direction and t represents time. The selection of the scanning frequency needs to comprehensively consider the belt speed and the rate of change of the material flow. A higher scanning frequency can obtain more detailed material flow morphology data, but it will increase the data processing burden. Step S12 involves installing a high-speed industrial camera adjacent to the 3D laser contour scanning unit and acquiring iron ore surface texture images at a preset frame rate to assist in judging the particle size segregation. By analyzing the surface texture roughness and particle distribution characteristics through image processing algorithms, the particle size differences in different areas in the belt width direction can be inferred. Step S13 involves continuously blowing away dust from the optical window of the 3D laser contour scanning unit using an air knife purging device to prevent dust adhesion from affecting scanning accuracy. The dust concentration in the port iron ore transportation environment is high; without protective measures, the optical window of the laser scanner will be covered by dust, leading to decreased measurement accuracy or even measurement failure. The air knife purging device uses compressed air to create a high-speed airflow to blow away dust from the surface of the optical window, keeping the optical channel clean. Step S14 involves transmitting the acquired cross-sectional height contour function h(x,t) and surface texture image data to the intelligent sampling control terminal in real time, completing the acquisition of material flow cross-sectional morphology data. Data transmission can use industrial Ethernet, fieldbus, or wireless transmission methods; the transmission protocol must ensure the real-time performance and reliability of the data.

[0066] Step S2 involves calculating the instantaneous virtual mass flux. Based on the cross-sectional profile data, the real-time belt speed, and preset bulk density parameters, the instantaneous virtual mass flux at the current moment is calculated. The instantaneous virtual mass flux represents the mass of material passing through a specific cross-section of the belt per unit time and forms the basis for subsequent cumulative flux calculations. Since directly measuring instantaneous mass flow rate requires dynamic metering equipment and its accuracy is significantly affected by factors such as vibration, this invention employs an indirect calculation method based on geometric measurement. This method estimates the mass flow rate by measuring the cross-sectional area of ​​the material flow and combining it with the material bulk density and belt speed. This method fully utilizes the cross-sectional morphology data already acquired by the 3D laser profile scanning unit, eliminating the need for additional metering equipment.

[0067] The formula for calculating the instantaneous virtual mass flux Φ(t) is:

[0068] In the formula, Φ(t) is the instantaneous virtual mass flux at time t; ρ is the preset bulk density of iron ore; v(t) is the real-time running speed of the conveyor belt at time t; h(x,t) is the material layer height at position x in the width direction of the conveyor belt at time t; W is the effective width of the conveyor belt. The integral of the cross-sectional profile height function along the belt width direction represents the cross-sectional area of ​​the material flow at time t. The physical meaning of this formula is that the volumetric flow rate is obtained by multiplying the cross-sectional area of ​​the material flow by the belt speed, and the mass flow rate is obtained by multiplying the volumetric flow rate by the bulk density of the material. The bulk density ρ is a preset parameter that can be set according to the iron ore type and historical data, or it can be corrected through subsequent weighing feedback. The belt speed v(t) is measured in real time by the belt speed detection unit and can be measured using equipment such as encoders, speed radar, or photoelectric tachometers.

[0069] Step S3 involves calculating the material flow pattern dispersion and adjusting the dynamic sampling trigger threshold. The material flow pattern dispersion is calculated based on the cross-sectional profile data, and the preset sampling trigger threshold is adaptively compensated and adjusted based on this dispersion to obtain the dynamic sampling trigger threshold. Traditional sampling methods use fixed sampling intervals or fixed cumulative mass thresholds to trigger sampling, without considering the time-varying characteristics of the material flow pattern. When the material flow fluctuates drastically, fixed threshold sampling may lead to a mismatch between the sample collection timing and the changes in material flow characteristics; when the material flow is stable, overly sensitive sampling triggering increases the number of unnecessary samplings. This invention introduces a material flow pattern dispersion parameter to achieve adaptive adjustment of the sampling trigger threshold, allowing the sampling process to better adapt to the actual changes in the material flow.

[0070] Step S3 specifically includes the following sub-steps. Step S31 involves calculating the ratio of the standard deviation to the mean of the cross-sectional area sequence based on the cross-sectional profile data from multiple consecutive sampling times, thus obtaining the material flow pattern dispersion ΔS. The ratio of the standard deviation to the mean, i.e., the coefficient of variation, is a dimensionless index that measures the degree of data dispersion. It can eliminate the influence of the absolute value of the data and facilitate comparison under different working conditions. The larger the material flow pattern dispersion ΔS, the more drastic the fluctuation of the material flow within the observation time window; the smaller ΔS, the more stable and uniform the material flow. Step S32 involves setting a preset sampling trigger threshold Q based on the material flow pattern dispersion ΔS. set Adaptive compensation adjustment is performed. Dynamic sampling trigger threshold Q. dynamic The formula for calculating Q is: dynamic =Q set ×(1+α×ΔS);

[0071] In the formula, Q dynamic The threshold for dynamic sampling; Q setThe preset sampling trigger threshold is defined as follows: α is the dispersion compensation coefficient, preferably ranging from 0.1 to 0.5, which can be adjusted according to the actual application scenario and sampling accuracy requirements; ΔS is the material flow pattern dispersion, a dimensionless parameter. Step S33 describes the adjustment mechanism of this formula. When the material flow pattern dispersion ΔS is large, the value of the product α×ΔS increases, leading to a decrease in the dynamic sampling trigger threshold Q. dynamic The corresponding increase in the system delays the sampling trigger to wait for more material to pass through, thereby capturing the fluctuation characteristics of the material flow over a longer time window and improving the representativeness of the sample to material flow changes. When the material flow pattern dispersion ΔS is small, the value of the product α×ΔS approaches zero, and the dynamic sampling trigger threshold Q... dynamic Approaching the preset sampling trigger threshold Q set The system samples at a regular pace. This adaptive adjustment mechanism matches the sampling frequency with the material flow characteristics, optimizing sampling efficiency while ensuring sample representativeness.

[0072] Step S4 involves sampling trigger control based on cumulative flux. The instantaneous virtual mass flux is integrated over time to obtain the cumulative flux. When the cumulative flux reaches the dynamic sampling trigger threshold, a sampling trigger signal is sent to the sampler. Compared to traditional timed sampling, the sampling trigger method based on cumulative flux ensures that the material batch represented by each sample is basically consistent, avoiding uneven sampling intervals caused by fluctuations in material flow rate. This method complies with the national standard requirement that the sampling quantity be proportional to the material batch size.

[0073] Step S4 specifically includes the following sub-steps. Step S41 is the time t when the previous sampling is completed. last Starting with the instantaneous virtual mass flux Φ(t), we perform continuous-time integration to obtain the cumulative flux M(t). The formula for calculating the cumulative flux M(t) is:

[0074]

[0075] In the formula, M(t) is the cumulative flux at the current time t; t last The time when the last sampling was completed is τ; Φ(τ) is the instantaneous virtual mass flux at time τ, which is the integral variable. This integral calculation can be implemented using numerical integration methods in a practical system, such as the trapezoidal rule or Simpson's rule, with the integration step size matched to the scanning frequency of the 3D laser contour scanning unit. Step S42 involves real-time comparison of the cumulative flux M(t) with the dynamic sampling trigger threshold Q. dynamic The size of the flux. After the intelligent sampling control terminal completes the calculation of the instantaneous virtual mass flux in each scan cycle, it immediately updates the cumulative flux and compares it with the threshold. Step S43 is when M(t)≥Q dynamicAt the same time, the intelligent sampling control terminal sends a sampling trigger signal to the frequency converter driven sampler, controlling the sampler to perform the sampling action, and simultaneously resets the cumulative flux M(t) to zero and updates t. last To prepare for the next sampling cycle at the current moment, a dynamic sampling trigger based on virtual quality integral is completed.

[0076] Step S5 involves variable-speed cutting sampling based on cross-sectional thickness distribution. The material layer thickness distribution along the belt width is obtained based on the cross-sectional profile data. The lateral movement speed of the sampling head is dynamically adjusted based on this distribution, allowing the sampling head to cut at different speeds in different material layer thickness ranges. During belt conveying of iron ore, due to the combined effects of the material's inherent characteristics and conveyor vibration, the material layer typically exhibits an uneven distribution along the belt width, often showing an off-center load or a higher center and lower sides. Traditional sampling machines use a constant speed to cross the belt for cutting. When the sampling head passes through areas with thicker material layers, the amount of material to be cut per unit time increases, potentially causing the sampling hopper to fill instantly and overflow, affecting the sample's particle size representativeness. This invention adjusts the lateral movement speed of the sampling head according to the material layer thickness distribution. In areas with thicker material layers, the sampling head cuts at a lower speed, ensuring the sampling hopper has sufficient time to receive the entire layer of material in that area; in areas with thinner material layers, it cuts at a higher speed, improving sampling efficiency.

[0077] Step S5 specifically includes the following sub-steps. Step S51 involves dividing the belt width direction into N equally spaced intervals, and calculating the average material layer thickness h(y) of each interval and the average material layer thickness H of the entire cross-section based on the cross-sectional profile data. avg The selection of the number of intervals N needs to balance control accuracy and computational efficiency. Too few intervals will result in coarse speed regulation, while too many intervals will increase the complexity of the control system. h(y) represents the average material layer thickness of the interval at position y, H avg This is the arithmetic mean of the material layer thickness in all sections of the entire cross-section. Step S52 involves calculating the lateral velocity v of the sampling head in each section based on the material layer thickness h(y). cutter (y).

[0078] Lateral velocity v cutter The formula for calculating (y) is: ;

[0079] In the formula, v cutter (y) represents the lateral velocity of the sampling head at position y of the belt width; v base The reference lateral speed can be set according to the sampler design parameters and process requirements; H avgHere, h(y) represents the average thickness of the material layer across the entire cross-section; h(y) represents the thickness of the material layer at position y; ε is a zero-division constant, preferably ranging from 0.001 to 0.01 m, used to avoid division errors or abnormal calculation results when the material layer thickness h(y) is zero or extremely small; β is the damping coefficient, preferably ranging from 0.3 to 0.5, used to adjust the sensitivity of speed to thickness changes, with a larger β value resulting in a greater speed adjustment range. The control logic of this formula is that when the material layer thickness h(y) at a certain position is greater than the average thickness H... avg When, the ratio Less than 1, the lateral velocity v after the beta power operation cutter (y) is lower than the reference speed v base When the material layer thickness h(y) is less than the average thickness H avg At that time, the lateral velocity v cutter (y) is higher than the reference speed v base This inverse proportional adjustment relationship ensures that the sampling head slows down in areas with thicker material layers to fully absorb the material, and accelerates in areas with thinner material layers to improve efficiency. Step S53 involves the intelligent sampling control terminal adjusting the lateral speed v during the sampling head's traverse of the conveyor belt. cutter (y) Real-time control of the motor speed of the variable frequency drive sampler, so that the sampling head can cut at a lower speed in areas with a larger material layer thickness to fully accept the material, and cut at a higher speed in areas with a smaller material layer thickness, thus completing variable speed cutting sampling based on the cross-sectional thickness distribution.

[0080] Step S5 also includes a fault-tolerant protection mechanism. Step S54 involves processing the calculated lateral velocity v. cutter (y) performs boundary constraints, when v cutter (y) is greater than the maximum allowable lateral speed of the motor, v. max At that time, v cutter (y) is restricted to v max When v cutter (y) is less than the minimum lateral velocity v min At that time, v cutter (y) is restricted to v min The maximum and minimum speeds of the motor are limited by the mechanical structure and electrical parameters. Speed ​​boundary limits prevent the system from issuing control commands beyond the physical limits of the equipment, protecting its safe operation. Step S55 involves automatically performing two cuts to sample the area when the thickness of a local section of the material layer is too large, making it difficult to completely capture the entire layer in a single cut. This ensures the completeness and representativeness of the sample. The system determines whether multiple cuts are necessary by pre-calculating the maximum amount of material that can be accepted in a single cut and comparing it with the predicted amount of material in the area.

[0081] Step S6 involves weighing feedback and parameter correction optimization of the sample quantity. The weighing feedback unit verifies the weight of a single sample and feeds the weighing data back to the intelligent sampling control terminal, thereby optimizing the sampling parameters. The calculation of instantaneous virtual mass flux relies on preset bulk density parameters, while the actual bulk density of iron ore varies due to factors such as variety, moisture content, and particle size distribution. By comparing the measured sample quantity with the theoretically calculated value, the accuracy of the bulk density parameters can be assessed and corrected, forming a closed-loop control system and continuously improving the system's measurement accuracy.

[0082] Step S6 specifically includes the following sub-steps. Step S61 involves setting a weighing sensor inside the sampling hopper or setting an electronic belt scale on the sampling conveying path to weigh the sample obtained each time in real time, thereby obtaining the measured sampling quantity M. real The weighing sensor can be of strain gauge, capacitive, or piezoelectric type, and the installation location and measuring range need to be selected according to the sampler structure and sampling range. Step S62 is to input the measured sampling quantity M real The sampling deviation rate is calculated by comparing the measured sample size with the theoretical sample size. The theoretical sample size can be estimated based on the instantaneous virtual mass flux within a certain period before and after the sampling trigger. The sampling deviation rate represents the relative deviation between the measured value and the theoretical value. Step S63 involves feeding the sampling deviation rate back to the intelligent sampling control terminal. When the sampling deviation rate exceeds a preset threshold, the preset bulk density parameter ρ or the dispersion compensation coefficient α is automatically adjusted to complete the closed-loop correction and optimization of the sampling parameters. If the measured sample size is consistently higher than the theoretical value, it indicates that the actual bulk density is greater than the preset value, and the system accordingly increases the bulk density parameter; otherwise, it decreases it. This adaptive correction mechanism ensures that the system parameters continuously approximate actual operating conditions, improving the accuracy and stability of long-term operation.

[0083] This invention also provides an iron ore conveyor belt sampling system for implementing the aforementioned iron ore conveyor belt sampling method. The system includes a 3D laser contour scanning unit, a surface texture imaging unit, an intelligent sampling control terminal, a frequency converter-driven sampler, a weighing feedback unit, and a conveyor belt speed detection unit. These units are connected via signal cables or an industrial communication network to form a complete control closed loop encompassing data acquisition, analysis and processing, control execution, and feedback correction.

[0084] The 3D laser profile scanning unit is installed upstream of the sampling point on the belt conveyor to acquire real-time cross-sectional profile data of the iron ore flow. This unit employs a combination of a line laser and a high-speed image sensor. The line laser emits a linear laser line projected onto the surface of the flow, while the image sensor captures images of the laser line's deformation on the surface from the side. The height coordinates of each point on the flow surface are calculated using triangulation principles. The 3D laser profile scanning unit is installed 5 to 10 meters upstream of the sampling point on the belt conveyor. This installation distance allows the system to complete the acquisition and processing of the cross-sectional shape data of the flow before the sampling action, providing sufficient data preparation time for sampling decisions. The 3D laser profile scanning unit is equipped with an air knife purging device, which consists of a compressed air source and a flat nozzle. The nozzle faces the optical window of the scanner, continuously spraying a high-speed airflow to form an air curtain, effectively blocking and removing dust particles from the environment, keeping the optical window clean, and ensuring that measurement accuracy is not affected by dust.

[0085] The surface texture imaging unit is installed adjacent to the 3D laser contour scanning unit to acquire surface texture image data of iron ore. The surface texture imaging unit is a high-speed industrial camera, employing either a CCD or CMOS image sensor, and works in conjunction with a ring-shaped LED illumination device to obtain clear images of the material flow surface. The surface texture imaging unit assists in determining the particle size segregation of the iron ore. By analyzing the surface roughness distribution through image processing algorithms, it can infer the particle size differences in different regions along the belt width, providing image-based evidence for calculating the transverse segregation index.

[0086] The intelligent sampling control terminal is connected to both the 3D laser contour scanning unit and the surface texture imaging unit, receiving cross-sectional contour data and surface texture image data. It is responsible for analyzing and processing material flow morphology data and generating sampling control decisions. The intelligent sampling control terminal can be an industrial control computer or an embedded controller, with built-in dedicated data processing software and control algorithms. The intelligent sampling control terminal includes an image processing module and a sampling decision module. These two modules can be independent hardware units or software functional modules running on the same processor. The image processing module processes the cross-sectional contour data and surface texture image data, performing data filtering, feature extraction, and parameter calculation to extract material flow morphology characteristic parameters, including cross-sectional area, material layer thickness distribution, and material flow morphology dispersion. The sampling decision module calculates the instantaneous virtual mass flux and cumulative flux based on the material flow morphology characteristic parameters, determines whether the dynamic sampling trigger threshold has been reached, calculates the dynamic lateral velocity curve of the sampling head, and generates corresponding control commands to send to the frequency converter-driven sampler.

[0087] The variable frequency drive (VFD) sampler is connected to an intelligent sampling control terminal, receiving sampling trigger signals and dynamic lateral speed commands to execute variable speed cutting sampling actions. The VFD sampler includes a sampling head and a VFD drive motor. The sampling head is the cutting component that directly contacts the material; its opening size is no less than three times the maximum particle size of the iron ore. This size requirement complies with national standards regarding sampler opening sizes, ensuring that materials of various particle sizes can enter the sampling hopper without obstruction, avoiding sampling deviations caused by large particles clogging the sampling port. The VFD drive motor is connected to the intelligent sampling control terminal, adjusting the motor speed according to the dynamic lateral speed commands, thereby changing the lateral speed of the sampling head and achieving variable speed cutting during the sampling head's traverse of the conveyor belt. The VFD drive motor can be an AC asynchronous motor with a frequency converter, or a servo motor with a servo driver; the latter provides higher speed control accuracy and faster response speed.

[0088] The weighing feedback unit is connected to the intelligent sampling control terminal to verify the sampling quantity by weighing and to provide feedback on the weighing data. The weighing feedback unit includes a weighing sensor installed inside the sampling hopper or an electronic belt scale installed on the sampling conveying path. The built-in weighing sensor in the sampling hopper can directly measure the mass of each sample taken, while the electronic belt scale can measure the cumulative mass of the sample during transport. The weighing feedback unit acquires the single sampling quantity in real time and feeds the weighing data back to the intelligent sampling control terminal. The intelligent sampling control terminal compares the measured sampling quantity with the theoretically calculated value, and accordingly corrects the bulk density parameter and other control parameters, achieving closed-loop control of the sampling system.

[0089] The system also includes a belt speed detection unit, which is connected to the intelligent sampling control terminal to acquire the belt speed in real time. The belt speed detection unit can employ a contact encoder or a non-contact laser tachometer. The encoder measures the rotational speed and converts it into linear velocity by coaxially connecting to the driven pulley or idler of the belt. The laser tachometer measures the moving speed of the belt surface through the Doppler effect. Belt speed is a key parameter in the instantaneous virtual mass flux calculation formula, and its measurement accuracy directly affects the accuracy of mass flow estimation. Therefore, the belt speed detection unit should be selected from speed measuring devices with appropriate accuracy and fast response.

[0090] In the iron ore conveyor belt sampling system of this invention, a 3D laser contour scanning unit and a surface texture imaging unit serve as front-end sensing devices to acquire raw data on the material flow morphology. An intelligent sampling control terminal acts as the data processing and decision-making center, analyzing the raw data and generating control commands. A variable frequency drive sampler acts as the actuator, completing the sampling action according to the control commands. A weighing feedback unit and a conveyor belt speed detection unit serve as feedback elements, providing the measured data required for closed-loop correction of the control system. All units work collaboratively to form a complete control loop of perception, decision-making, execution, and feedback, achieving intelligent and adaptive control of the iron ore conveyor belt sampling process.

[0091] Example 1: This example uses a belt conveyor sampling scenario at a port iron ore terminal to illustrate the iron ore belt sampling method and system of the present invention.

[0092] The iron ore conveyor belt sampling system used in this embodiment includes a 3D laser contour scanning unit, a surface texture imaging unit, an intelligent sampling control terminal, a frequency converter-driven sampler, a weighing feedback unit, and a belt speed detection unit. The effective width w of the belt conveyor is 1.4m, the designed conveying capacity is 3000t / h, and the belt running speed is 2.5m / s. The 3D laser contour scanning unit is installed 8m upstream of the sampling point, using a line laser in conjunction with a high-speed CMOS sensor, with a scanning frequency set to 200Hz and a measurement accuracy of ±1mm. The surface texture imaging unit uses an industrial camera with a resolution of 1280×1024 and a frame rate set to 30fps. The intelligent sampling control terminal uses an industrial control computer. The sampling head opening width of the frequency converter-driven sampler is 150mm, the opening height is 200mm, and the sampling head opening size is more than three times the maximum iron ore particle size of 50mm, conforming to the requirements of GB / T10322.1-2014. The weighing feedback unit uses a weighing sensor with a range of 50 kg and an accuracy of ±0.1 kg. The iron ore conveyor belt sampling method in this embodiment is implemented according to the following steps:

[0093] Step S1 involves acquiring the cross-sectional morphology data of the material flow. In step S11, a 3D laser contour scanning unit is installed 8m upstream of the sampling point on the belt conveyor. It scans the cross-sectional contour of the iron ore flow on the belt in real time at a scanning frequency of 200Hz, acquiring the material layer height values ​​at 1400 locations along the belt width, forming the cross-sectional height contour function h(x,t). In step S12, a high-speed industrial camera is installed adjacent to the 3D laser contour scanning unit, acquiring surface texture images of the iron ore at a frame rate of 30fps. In step S13, an air knife purging device continuously purifies the optical window of the 3D laser contour scanning unit with compressed air at 0.5MPa. In step S14, the cross-sectional height contour function h(x,t) and surface texture image data are transmitted in real time to the intelligent sampling control terminal via industrial Ethernet.

[0094] In step S2, the preset bulk density ρ of iron ore is set to 2100 kg / m³. 3 The belt speed detection unit measures the belt speed v(t) in real time, and the intelligent sampling control terminal uses the instantaneous virtual mass flux formula:

[0095] Calculate the instantaneous virtual mass flux at the current moment.

[0096] In step S3, the dispersion compensation coefficient α is set to 0.3, and the preset sampling trigger threshold Q is...set The weight is set to 500 kg. In step S31, the intelligent sampling control terminal calculates the ratio of the standard deviation to the mean of the cross-sectional area sequence based on the cross-sectional profile data from 50 consecutive sampling times, thus obtaining the material flow pattern dispersion ΔS. In step S32, the dynamic sampling trigger threshold formula Q is used... dynamic =Q set The dynamic sampling trigger threshold is calculated using the formula ×(1+α×ΔS). In step S33, when the material flow pattern dispersion ΔS is 0.15, the dynamic sampling trigger threshold Q is... dynamic The calculation is 500×(1+0.3×0.15)=522.5kg.

[0097] In step S4, step S41 refers to the time t when the previous sampling was completed. last Starting from the instantaneous virtual mass flux Φ(t), the cumulative flux M(t) is obtained by continuous-time integration. Step S42 compares the cumulative flux M(t) with the dynamic sampling trigger threshold Q in real time. dynamic The size of M(t). In step S43, when M(t) ≥ Q dynamic At that time, the intelligent sampling control terminal sends a sampling trigger signal to the frequency converter driven sampler.

[0098] In step S5, step S51 divides the belt width direction into 14 equally spaced intervals, and calculates the average material layer thickness h(y) of each interval and the average material layer thickness H of the entire cross section. avg In step S52, the reference lateral velocity v base The velocity is set to 0.6 m / s, the zero constant ε is set to 0.005 m, and the damping coefficient β is set to 0.4. Based on the lateral velocity formula v... cutter (y)=v base × Calculate the lateral velocity of the sampling head in each interval. When the material layer thickness h(y) in a certain interval is 0.12m and the average material layer thickness H across the entire cross-section is... avg When the distance is 0.08m, the lateral velocity v in this section is... cutter (y) is calculated as m / s. In step S53, the intelligent sampling control terminal determines the lateral speed v. cutter (y) Real-time control of the variable frequency drive motor speed. In step S54, the maximum lateral speed v max The minimum lateral speed is set to 0.9 m / s. min The speed is set to 0.3 m / s. In step S55, when the predicted material amount for a single cut exceeds 80% of the sampling hopper capacity, the system automatically performs two cuts.

[0099] In step S6, step S61 involves using a weighing sensor to weigh the sample taken each time in real time to obtain the measured sample quantity M. real Step S62 will measure the actual sample quantity M.real The sampling deviation rate is calculated by comparing the sample size with the theoretical sample size. In step S63, when the sampling deviation rate exceeds 5%, the preset bulk density parameter ρ is automatically adjusted.

[0100] Example 2 uses a large imported iron ore unloading terminal as an application scenario. The effective width W of the belt conveyor is 2.0m, the designed conveying capacity is 6000t / h, and the belt running speed is 4.0m / s. The 3D laser contour scanning unit is installed 10m upstream of the sampling point, and the scanning frequency is set to 500Hz.

[0101] The main differences in parameter settings between this embodiment and Embodiment 1 are as follows: The dispersion compensation coefficient α is set to 0.5 to adapt to the large fluctuations in material flow at large wharves. The preset sampling trigger threshold Q is... set The load is set to 800 kg. The zero constant ε is set to 0.01 m, and the damping coefficient β is set to 0.5. The reference lateral velocity v... base The maximum lateral speed is set to 0.8 m / s. max The minimum lateral speed is set to 1.2 m / s. min The velocity is set to 0.4 m / s. The initial bulk density ρ is set to 2200 kg / m³. 3 .

[0102] This embodiment uses the same steps S1 to S6 as Embodiment 1 for sampling operations. In step S3, when the material flow pattern dispersion ΔS is 0.25, the dynamic sampling trigger threshold Q is set. dynamic The calculated weight is 800 × (1 + 0.5 × 0.25) = 900 kg.

[0103] Example 3: This example uses a belt conveyor sampling scenario in a mineral processing plant. The effective width W of the belt conveyor is 0.8m, the designed conveying capacity is 800t / h, and the belt speed is 1.5m / s. The 3D laser contour scanning unit is installed 5m upstream of the sampling point, and the scanning frequency is set to 100Hz.

[0104] The main differences in parameter settings between this embodiment and Embodiment 1 are as follows: The dispersion compensation coefficient α is set to 0.1 to adapt to the relatively stable material flow conditions in the mineral processing plant. The preset sampling trigger threshold Q is... set The load is set to 200 kg. The zero constant ε is set to 0.001 m, and the damping coefficient β is set to 0.3. The reference lateral velocity v... base The maximum lateral speed is set to 0.4 m / s. max The minimum lateral speed is set to 0.6 m / s. min The velocity is set to 0.2 m / s. The initial bulk density ρ is set to 1900 kg / m³. 3 This embodiment uses the same steps S1 to S6 as in Embodiment 1 for sampling.

[0105] Comparative Example 1 employs a timed sampling method. The hardware configuration of the sampling system is the same as in Example 1, but the sampling triggering method differs. Comparative Example 1 does not use a dynamic sampling triggering strategy based on virtual mass integrals; instead, it uses a fixed time interval to trigger sampling. Based on the conveying capacity and preset sampling triggering threshold of Example 1, the average sampling interval is calculated to be approximately 10 seconds. Comparative Example 1 sets sampling to be triggered every 10 seconds. The transverse movement speed of the sampling head uses the same variable-speed cutting method as in Example 1.

[0106] Comparative Example 2 uses a uniform speed cutting method. The hardware configuration of the sampling system is the same as in Example 1, and the sampling triggering method adopts the same dynamic sampling triggering strategy based on virtual mass integral as in Example 1. However, the lateral movement method of the sampling head is different. Comparative Example 2 does not use variable speed cutting based on cross-sectional thickness distribution. The sampling head crosses the belt at a constant speed of 0.6 m / s to make cuts.

[0107] Comparative Example 3 employs a traditional timed sampling and uniform-speed cutting method, with the same hardware configuration as Example 1. Comparative Example 3 is set to trigger sampling every 10 seconds, with the sampling head moving across the belt at a constant speed of 0.6 m / s to make cuts. Comparative Example 3 represents the operating mode of a conventional belt sampling system in the prior art.

[0108] Comparative Experiment: To verify the implementation results of Examples 1, 2, and 3 compared with Comparative Examples 1, 2, and 3, the following comparative experiment was conducted. The experiment was carried out in accordance with the relevant provisions of GB / T10322.1-2014 Iron Ore Sampling and Sample Preparation Methods and ISO3082:2017 Iron Ore Sampling and Sample Preparation Procedures.

[0109] The experiment used imported Brazilian iron ore from a certain port as the experimental material. The iron ore particle size ranged from 0 to 50 mm, and the iron content ranged from 62% to 65%. The experiment was conducted under fluctuating material flow conditions, with the material flow thickness artificially set to fluctuate randomly between 0.05 m and 0.15 m, and the material flow eccentricity varying between 0% and 30%. Each sampling method was used to take 50 consecutive samples, and the sample size, material flow state at the time of sampling, and particle size distribution of the sample were recorded for each sampling.

[0110] The stability of the sampling quantity is evaluated using the coefficient of variation (CV). The CV is equal to the ratio of the standard deviation of the sampling quantity to the mean of the sampling quantity. The smaller the CV, the more stable the sampling quantity. GB / T10322.1-2014 stipulates that the coefficient of variation of mechanical sampling should not exceed 20%.

[0111] Particle size representativeness was evaluated using the Particle Size Deviation Index (PDI), which is equal to the root mean square deviation between the sample particle size distribution and the actual particle size distribution of the material flow. The PDI was calculated by sieving and analyzing the samples after each sampling, comparing the results with the actual particle size distribution of the material flow during the same period. A smaller PDI indicates better particle size representativeness of the sample.

[0112] The sampling deviation rate is evaluated using the absolute value of the difference between the iron grade of the sample and the iron grade of the actual material flow. The experiment is conducted by chemically analyzing the iron grade of a mixed sample of 50 samples taken in each group, and comparing it with the actual iron grade of the same batch of material flow.

[0113] Table 1. Results of the experimental comparison on the stability of sampling amount

[0114]

[0115] Table 1 presents the results of the sampling stability comparison experiment for each experimental group. The sample mean represents the average sample size from 50 samplings. The sample standard deviation represents the dispersion of the 50 samplings. The coefficient of variation (CV) represents the relative dispersion of the sample size, calculated by dividing the sample standard deviation by the sample mean and then multiplying by 100%.

[0116] As shown in Table 1, the coefficient of variation (COP) for Example 1 was 7.98%, for Example 2 it was 8.99%, and for Example 3 it was 6.94%, all meeting the requirement of GB / T10322.1-2014 that the COP should not exceed 20%. The COP for Comparative Example 1 was 28.03%, and for Comparative Example 3 it was 28.82%, both exceeding the standard requirements. The COP for Comparative Example 2 was 8.49%, close to that of Example 1.

[0117] Compared to Comparative Examples 1 and 3, Examples 1, 2, and 3 show a reduction in the coefficient of variation of approximately 20 percentage points. The examples employ a dynamic sampling triggering strategy based on virtual mass integrals. When the material flow rate is high, the sampling interval is shortened; when the material flow rate is low, the sampling interval is extended, ensuring that the material batch represented by each sample is essentially consistent. Comparative Examples 1 and 3 use fixed time intervals for sampling, resulting in drastic fluctuations in the sample size due to variations in material flow thickness under fluctuating material flow conditions. Comparative Example 2 employs a dynamic sampling triggering strategy, therefore its coefficient of variation is close to that of the examples, indicating that the dynamic sampling triggering strategy of this invention plays a crucial role in improving the stability of the sample size.

[0118] Table 2. Results of the representative particle size comparison experiment.

[0119]

[0120] Table 2 presents the results of the particle size representativeness comparison experiment for each experimental group. The Particle Size Deviation Index (PDI) is the root mean square value of the deviation of each particle size distribution between the sample particle size distribution and the actual material flow particle size distribution. The deviation of the particle size distribution greater than 25 mm represents the absolute value of the difference between the content of particles greater than 25 mm in the sample and the content of that particle size in the actual material flow. The deviation of the particle size distribution less than 5 mm represents the absolute value of the difference between the content of particles less than 5 mm in the sample and the content of that particle size in the actual material flow. The deviation of each particle size distribution was determined by sieving test according to the method in Appendix A of GB / T10322.1-2014.

[0121] As shown in Table 2, the particle size deviation index of Example 1 was 2.13%, that of Example 2 was 2.56%, and that of Example 3 was 1.87%. The particle size deviation index of Comparative Example 1 was 3.45%, that of Comparative Example 2 was 5.82%, and that of Comparative Example 3 was 6.17%.

[0122] Compared to Comparative Examples 2 and 3, the particle size deviation index of Example 1 was reduced by approximately 3 to 4 percentage points. Example 1 employed a variable-speed cutting sampling method based on cross-sectional thickness distribution. When the sampling head passed through a thicker material layer, the lateral speed was reduced to fully accommodate the material and avoid particle size deviation caused by overflow. During belt conveying of iron ore, large particles tend to roll to the sides of the belt, forming an off-center load. The material layer in the off-center load area is thicker, and traditional uniform-speed cutting methods are prone to overflow in this area, missing large particles and resulting in a lower content of large particles in the sample. Example 1 effectively solved this problem through variable-speed cutting. Although the particle size deviation index of Comparative Example 1 was lower than that of Comparative Examples 2 and 3, the statistical averaging effect of a large number of samples after multiple samplings in Comparative Example 1 partially offset the particle size deviation of a single sampling. However, its excessively large coefficient of variation, leading to unstable sampling volume, still exists.

[0123] Table 3. Comparison of Sampling Deviation Rate Experimental Results

[0124]

[0125] Table 3 presents the comparative experimental results of sampling deviation rates for each experimental group. The sample iron grade is the total iron content determined by chemical analysis of a mixed sample from 50 samplings. The actual iron grade is the total iron content determined by full-volume sampling of the same batch of material. The sampling deviation rate is the ratio of the absolute value of the difference between the sample iron grade and the actual iron grade to the actual iron grade. It is calculated as: sampling deviation rate = sample iron grade minus the absolute value of the actual iron grade, divided by the actual iron grade, and then multiplied by 100%. The iron grade determination method follows GB / T6730.5-2007, "Determination of Total Iron Content in Iron Ore - Titanium Trichloride Reduction of Potassium Dichromate Titration Method".

[0126] As can be seen from Table 3, the sampling deviation rate of Example 1 was 0.25%, that of Example 2 was 0.26%, and that of Example 3 was 0.17%. The sampling deviation rate of Comparative Example 1 was 0.58%, that of Comparative Example 2 was 1.15%, and that of Comparative Example 3 was 1.43%.

[0127] Compared to the comparative examples, the sampling deviation rate of the examples was reduced by approximately 0.3 to 1.2 percentage points. According to ISO 3082:2017, the precision requirement for iron ore trading sampling is an iron grade deviation of no more than 0.5%. The sampling deviation rates of Examples 1, 2, and 3 all meet this requirement, while the sampling deviation rates of Comparative Examples 2 and 3 exceed this requirement. The reduction in the sampling deviation rate is the result of the synergistic effect of the dynamic sampling triggering strategy and the variable-speed cutting sampling method of this invention. The dynamic sampling triggering strategy ensures that the sampling timing matches the material flow state, while the variable-speed cutting sampling method ensures the particle size representativeness of the sample. Together, they improve the overall representativeness of the sample for the material batch and reduce sampling deviation.

[0128] Table 4 Comparison of coefficients of variation under different material flow fluctuation conditions

[0129]

[0130] Table 4 shows the comparison of the coefficients of variation between Example 1 and Comparative Example 3 under different material flow fluctuation conditions. For the mild fluctuation condition, the material flow thickness was set to vary between 0.07 m and 0.09 m, with a fluctuation range of approximately 25%. For the moderate fluctuation condition, the material flow thickness was set to vary between 0.05 m and 0.15 m, with a fluctuation range of approximately 100%, the same as the experimental conditions in Tables 1 to 3. For the severe fluctuation condition, the material flow thickness was set to vary between 0.03 m and 0.18 m, with a fluctuation range of approximately 200%. The decrease in the coefficient of variation is calculated by subtracting the coefficient of variation of Example 1 from the coefficient of variation of Comparative Example 3.

[0131] As shown in Table 4, with the increase in the degree of material flow fluctuation, the coefficient of variation of Example 1 increased from 5.23% to 11.45%, and the coefficient of variation of Comparative Example 3 increased from 15.67% to 42.36%. The coefficient of variation of Example 1 under various fluctuation conditions met the requirements of GB / T10322.1-2014, while the coefficient of variation of Comparative Example 3 under moderate and severe fluctuation conditions exceeded the standard requirements. The decrease in coefficient of variation increased with the degree of fluctuation, reaching 30.91 percentage points under severe fluctuation conditions. The dynamic sampling triggering strategy of the present invention is more effective under conditions of severe material flow fluctuation. It senses the degree of material flow fluctuation through the material flow morphology dispersion parameter and adaptively adjusts the sampling triggering threshold. When the fluctuation is severe, the sampling interval is increased to capture material flow characteristics within a longer time window, thereby effectively suppressing the fluctuation of the sampling amount.

[0132] Table 5 Comparison of Particle Size Deviation Index under Different Off-center Loading Degrees

[0133]

[0134] Table 5 shows the comparison results of particle size deviation indices for Example 1 and Comparative Example 2 under different degrees of off-center loading. Under no off-center loading, the material layer thickness was set to be uniformly distributed along the belt width. Under slight off-center loading, the material layer thickness on one side of the belt was set to be 30% higher than the other side. Under moderate off-center loading, the material layer thickness on one side of the belt was set to be 60% higher than the other side, the same as the experimental conditions in Tables 1 to 3. Under severe off-center loading, the material layer thickness on one side of the belt was set to be 100% higher than the other side. The reduction in the deviation index was the particle size deviation index of Comparative Example 2 minus the particle size deviation index of Example 1.

[0135] As shown in Table 5, with the increase of the degree of off-center loading, the particle size deviation index of Example 1 increased from 1.45% to 2.78%, and the particle size deviation index of Comparative Example 2 increased from 2.18% to 8.56%. The decrease in the deviation index increased with the increase of the degree of off-center loading, reaching 5.78 percentage points under heavy off-center loading conditions. This indicates that the variable speed cutting sampling method of the present invention is more effective under off-center loading conditions. In the off-center loading area, the material layer is thicker and enriched with large particles. In the traditional uniform speed cutting method, the transverse cutting speed in this area is too fast, causing the sampling bucket to be unable to accept the entire layer of material in time, resulting in overflow. The missed material is mainly large particles, resulting in a low content of large particles in the sample. The present invention reduces the transverse speed in the thicker material layer area to ensure that the sampling bucket fully accepts the entire layer of material in this area, effectively avoiding particle size deviation caused by overflow.

[0136] Based on the comparative experimental results above, the iron ore conveyor belt sampling method of the present invention, employing a dynamic sampling triggering strategy based on virtual mass integrals, can reduce the sampling quantity variation coefficient by approximately 20 percentage points. Under fluctuating material flow conditions, the sampling quantity stability is significantly better than that of timed sampling methods, meeting the requirements of GB / T10322.1-2014 for the mechanical sampling mass variation coefficient. The variable-speed cutting sampling method based on cross-sectional thickness distribution can reduce the particle size deviation index by approximately 3 to 6 percentage points. Under eccentric loading conditions, the particle size representativeness is significantly better than that of uniform-speed cutting methods, effectively solving the problem of large particle under-collection caused by eccentric loading. The synergistic effect of the dynamic sampling triggering strategy and the variable-speed cutting sampling method can reduce the sampling deviation rate by approximately 0.3 to 1.2 percentage points, meeting the requirements of ISO3082:2017 for the precision of iron ore trade sampling.

[0137] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of the invention. All equivalent changes made based on the description and drawings of the present invention are included within the scope of the present invention.

Claims

1. A method for sampling iron ore conveyor belts, characterized in that, Includes the following steps: S1. By installing a 3D laser contour scanning unit and a surface texture imaging unit upstream of the sampling point of the belt conveyor, the cross-sectional contour data and surface texture image data of the iron ore flow are acquired in real time. S2. Based on the cross-sectional profile data, combined with the real-time running speed of the belt and the preset bulk density parameters, calculate the instantaneous virtual mass flux at the current moment. S3. Calculate the material flow pattern dispersion based on the cross-sectional profile data, and adaptively compensate and adjust the preset sampling trigger threshold based on the material flow pattern dispersion to obtain the dynamic sampling trigger threshold. S4. Integrate the instantaneous virtual mass flux over time to obtain the cumulative flux. When the cumulative flux reaches the dynamic sampling trigger threshold, send a sampling trigger signal to the sampler. S5. Obtain the material layer thickness distribution in the belt width direction based on the cross-sectional profile data, and dynamically adjust the transverse movement speed of the sampling head based on the material layer thickness distribution so that the sampling head can perform transverse cutting at different speeds in different material layer thickness areas. S6. The single sampling quantity is weighed and verified through the weighing feedback unit, and the weighing data is fed back to the intelligent sampling control terminal for the correction and optimization of sampling parameters.

2. The iron ore belt sampling method according to claim 1, characterized in that, Step S1 includes: S11. Install the 3D laser contour scanning unit 5 to 10 meters upstream of the sampling point of the belt conveyor, and scan the cross-sectional contour of the iron ore flow on the belt in real time at a preset scanning frequency to obtain the material layer height value at each position in the width direction of the belt, and form the cross-sectional height contour function h(x,t). S12. Install a high-speed industrial camera adjacent to the 3D laser contour scanning unit to acquire iron ore surface texture images at a preset frame rate to help determine the particle size segregation situation. S13. The optical window of the 3D laser contour scanning unit is continuously purged by an air knife purging device to prevent dust from adhering and affecting the scanning accuracy. S14. The acquired cross-sectional height profile function h(x,t) and surface texture image data are transmitted to the intelligent sampling control terminal in real time to complete the acquisition of material flow cross-sectional morphology data.

3. The iron ore belt sampling method according to claim 2, characterized in that, In step S2, the formula for calculating the instantaneous virtual mass flux Φ(t) is: ； In the formula, Φ(t) is the instantaneous virtual mass flux at time t (kg / s); ρ is the preset bulk density of iron ore (kg / m³). 3 v(t) is the real-time running speed of the belt at time t (m / s); h(x,t) is the material layer height at position x in the width direction of the belt at time t (m); W is the effective width of the belt (m). The integral of the cross-sectional profile height function along the belt width direction represents the cross-sectional area (m²). 2 ).

4. The iron ore belt sampling method according to claim 3, characterized in that, Step S3 includes: S31. Based on the cross-sectional profile data at multiple consecutive sampling times, calculate the ratio of the standard deviation to the mean of the cross-sectional area sequence to obtain the material flow pattern dispersion ΔS. S32. Based on the material flow pattern dispersion ΔS, set the preset sampling trigger threshold Q. set Adaptive compensation adjustment is performed, and the trigger threshold Q is dynamically sampled. dynamic The calculation formula is: Q dynamic =Q set ×(1+α×ΔS); In the formula, Q dynamic The dynamic sampling trigger threshold (kg); Q set The preset sampling trigger threshold (kg) is α, which is the dispersion compensation coefficient, ranging from 0.1 to 0.5; ΔS is the material flow pattern dispersion, a dimensionless parameter. S33. When the dispersion of material flow pattern ΔS increases, the dynamic sampling trigger threshold increases accordingly to capture the fluctuation characteristics of material flow; when the dispersion of material flow pattern ΔS decreases, the dynamic sampling trigger threshold approaches the preset sampling trigger threshold, thus completing the adaptive adjustment of the sampling trigger threshold based on the material flow pattern characteristics.

5. The iron ore belt sampling method according to claim 4, characterized in that, Step S4 includes: S41, the time t when the previous sampling was completed last Starting with the instantaneous virtual mass flux Φ(t), we perform continuous-time integration to obtain the cumulative flux M(t). The formula for calculating the cumulative flux M(t) is as follows: ； In the formula, M(t) is the cumulative flux (kg) at the current time t; t last τ is the time (s) when the last sampling was completed; Φ(τ) is the instantaneous virtual mass flux (kg / s) of the integral variable at time τ. S42. Real-time comparison of cumulative flux M(t) and dynamic sampling trigger threshold Q. dynamic Size; S43, When M(t) ≥ Q dynamic At that time, the intelligent sampling control terminal sends a sampling trigger signal to the frequency converter driven sampler, and simultaneously resets the cumulative flux M(t) to zero and updates t. last At the current moment, a dynamic sampling trigger based on virtual quality integral is completed.

6. The iron ore belt sampling method according to claim 5, characterized in that, Step S5 includes: S51. Divide the belt width direction into N equally spaced intervals, and calculate the average material layer thickness h(y) of each interval and the average material layer thickness H of the entire cross section based on the cross-sectional profile data. avg ; S52. Calculate the transverse velocity v of the sampling head in each interval based on the material layer thickness h(y). cutter (y), lateral velocity v cutter The formula for calculating (y) is: ； In the formula, v cutter (y) represents the lateral velocity (m / s) of the sampling head at position y of the belt width; v base The reference lateral velocity (m / s); H avg ε is the average material layer thickness across the entire cross section (m); h(y) is the material layer thickness at position y (m); ε is the zero constant for damping, ranging from 0.001 to 0.01 (m); β is the damping coefficient, ranging from 0.3 to 0.

5. S53. During the process of the sampling head crossing the belt, the intelligent sampling control terminal adjusts the lateral movement speed v. cutter (y) Real-time control of the motor speed of the variable frequency drive sampler, so that the sampling head can complete variable speed cutting and sampling based on the cross-sectional thickness distribution.

7. The iron ore belt sampling method according to claim 6, characterized in that, Step S5 also includes: S54. The calculated lateral velocity v cutter (y) performs boundary constraints, when v cutter (y) is greater than the maximum allowable lateral speed of the motor, v. max At that time, v cutter (y) is restricted to v max When v cutter (y) is less than the minimum lateral velocity v min At that time, v cutter (y) is restricted to v min ; S55. When a single cut is insufficient to completely capture the material in a localized section of a material layer, the system automatically performs two cuts to complete the sampling of the localized section of the material layer, ensuring the integrity and representativeness of the sampling.

8. The iron ore belt sampling method according to claim 7, characterized in that, Step S6 includes: S61. Install a weighing sensor inside the sampling hopper or an electronic belt scale on the sampling conveying path to weigh the sample obtained each time in real time, and obtain the measured sampling quantity M. real ; S62. The measured sample quantity M real Compare the sample size with the theoretical sample size and calculate the sampling deviation rate; S63. Feedback the sampling deviation rate to the intelligent sampling control terminal. When the sampling deviation rate exceeds the preset threshold, automatically adjust the preset bulk density parameter ρ or the dispersion compensation coefficient α to complete the closed-loop correction and optimization of the sampling parameters.

9. An iron ore conveyor belt sampling system for implementing the iron ore conveyor belt sampling method according to any one of claims 1 to 8, characterized in that, include: The 3D laser contour scanning unit is installed upstream of the sampling point of the belt conveyor to acquire cross-sectional contour data of the iron ore flow in real time. The surface texture imaging unit is installed adjacent to the 3D laser contour scanning unit to acquire surface texture image data of iron ore. The intelligent sampling control terminal is connected to the 3D laser contour scanning unit and the surface texture imaging unit respectively. It receives cross-sectional contour data and surface texture image data, calculates instantaneous virtual mass flux, cumulative flux and material flow pattern dispersion, and adaptively adjusts the sampling trigger threshold based on the material flow pattern dispersion, as well as calculates the dynamic lateral movement speed of the sampling head according to the cross-sectional thickness distribution. The variable frequency drive sampler is connected to the intelligent sampling control terminal, receives sampling trigger signals and dynamic transverse speed commands, and executes variable speed cutting sampling actions. The weighing feedback unit is connected to the intelligent sampling control terminal to verify the weighing quantity and feed back the weighing data.

10. The iron ore conveyor belt sampling system according to claim 9, characterized in that, The 3D laser contour scanning unit is installed 5-10 meters upstream of the sampling point on the belt conveyor to acquire cross-sectional shape data of the material flow before sampling. The 3D laser contour scanning unit is equipped with an air knife blowing device to continuously blow the optical window to prevent dust adhesion. The surface texture imaging unit is a high-speed industrial camera to help determine the particle size segregation of iron ore and provide image basis for calculating the transverse segregation index. The intelligent sampling control terminal includes an image processing module and a sampling decision module. The image processing module processes cross-sectional contour data and surface texture image data to extract material flow morphology feature parameters; the sampling decision module calculates the dynamic sampling trigger threshold and dynamic lateral movement speed based on the material flow morphology feature parameters; the variable frequency drive sampler includes a sampling head and a variable frequency drive motor; the opening size of the sampling head is not less than three times the maximum particle size of the iron ore; the variable frequency drive motor is connected to the intelligent sampling control terminal for control, and adjusts the lateral movement speed of the sampling head according to the dynamic lateral movement speed command to achieve variable speed cutting during the sampling head's traversal of the belt; the weighing feedback unit includes a weighing sensor installed in the sampling hopper or an electronic belt scale installed on the sampling conveyor path, which acquires the single sampling volume in real time and feeds the weighing data back to the intelligent sampling control terminal for closed-loop correction; it also includes a belt speed detection unit, which is connected to the intelligent sampling control terminal signal to acquire the belt running speed in real time, providing speed parameters for the calculation of instantaneous virtual mass flux.

Images