Intelligent ore pre-separation system and method based on air flow separation

Abstract

The application provides an ore intelligent pre-selection system and method based on airflow separation, which identifies the spatial profile features and surface scattering features of ore particles at the upstream of the free falling area of the raw ore stream, and then generates the trajectory deviation coefficient of the ore particles; the trajectory deviation coefficient and the local flow field parameters of the raw ore stream are used to predict the landing point trajectory partition of the ore particles, the airflow intervention instruction is generated according to the decision influence relationship of the trajectory deviation coefficient on the landing point trajectory partition; the airflow intervention instruction is responded to, the ore removing airflow area and the ore retaining airflow area during the separation of the ore particles are determined, the ore particles meeting the waste rock features are put into the ore removing airflow area; after the action of the airflow intervention, the landing point response signals of the ore removing airflow area and the ore retaining airflow area are obtained, and the airflow intervention instruction for the subsequent separation of the ore particles is updated according to the landing point response signals. The technical scheme provided by the application can effectively separate the ore particles under the collision interference between the ore particles.

Description

Technical Field

[0001] This application relates to the field of ore sorting technology, and more specifically, to an intelligent pre-sorting system and method for ores based on airflow separation. Background Technology

[0002] Ore sorting is a core link in the development and utilization of mineral resources. Its technological development is closely related to changes in mineral resource endowment and industrial demand upgrades. In the early days, most minerals were high-grade and easy-to-sort ores, and traditional technologies such as gravity sorting (such as jigging and shaking tables) could meet the needs. However, as easy-to-mine high-grade minerals gradually become depleted, the proportion of low-grade and associated complex ores continues to increase. Traditional technologies face bottlenecks such as low sorting efficiency and insufficient resource utilization. To meet this challenge, ore sorting technology is evolving towards refinement and intelligence, providing key technical support for the green and efficient utilization of mineral resources.

[0003] In existing ore sorting, ore sorting mainly relies on physicochemical differences. After the minerals are liberated through crushing and screening, gravity separation with density differences is used, such as jigging and shaking tables, to separate particles into layers. Magnetic separators collect iron and manganese minerals based on magnetic differences, while flotation uses collectors and air bubbles to select and adhere sulfide minerals. Photoelectric and X-ray identification is used to identify differences in reflectivity or atomic number, and high-speed air valves remove waste rock, achieving concentrate enrichment and tailings disposal. However, in intelligent ore sorting based on airflow separation, especially in the dense flow field formed by the free fall of ore particles, the mechanical collisions between ore particles and the obstruction and disturbance of subsequent airflow by particle groups cause a large number of ore particles to fail to be sorted into the corresponding zones, resulting in a decrease in the sorting accuracy of ore particles. Therefore, how to effectively sort ore particles under the interference of collisions between ore particles has become a challenge for the industry. Summary of the Invention

[0004] This application provides an intelligent pre-selection system and method for ores based on airflow separation, which can effectively separate ores under the interference of collisions between ores.

[0005] In a first aspect, this application provides a smart pre-selection method for ores based on airflow separation, comprising the following steps:

[0006] The raw ore stream is fed evenly into the sorting channel by a vibrating feeder, so that the raw ore stream forms a controllable sparse distribution in the free fall area;

[0007] The spatial contour features and surface scattering features of ore particles are identified upstream of the free fall zone, and the trajectory offset coefficient of the ore particles is generated based on the spatial contour features and the surface scattering features.

[0008] The trajectory deviation coefficient and the local flow field parameters of the original ore flow are used to predict the landing trajectory partitions of ore particles, and the landing trajectory partitions are mapped as ore sorting windows. Within the ore sorting window, airflow intervention commands are generated based on the decision influence relationship of the trajectory deviation coefficient on the landing trajectory partitions.

[0009] In response to the airflow intervention command, the ore rejection airflow zone and the ore retention airflow zone in the ore particle sorting process are determined, and the ore particles that meet the waste rock characteristics are entered into the ore rejection airflow zone by the multi-stage pulse air curtain, while the ore particles that do not meet the waste rock characteristics are entered into the ore retention airflow zone.

[0010] After the airflow intervention, the landing point response signal of the ore particles is obtained by the receiving end set in the ore rejection airflow zone and the ore retention airflow zone, and the airflow intervention command for subsequent ore particle sorting is updated with the landing point response signal.

[0011] In some embodiments, identifying the spatial profile features and surface scattering features of ore particles upstream of the free fall zone specifically includes:

[0012] Multi-view synchronous imaging of ore particles was performed upstream of the free fall zone using a high-speed linear array camera to obtain the original image dataset of the ore particles.

[0013] The original image dataset is preprocessed to obtain effective images of ore particles;

[0014] An edge detection algorithm is used to extract the spatial contour features of ore particles from the valid image;

[0015] Spectral analysis is performed on the effective image to obtain the scattered light intensity distribution data on the surface of the ore particles;

[0016] The surface scattering characteristics of the ore particles are determined based on the scattered light intensity distribution data.

[0017] In some embodiments, generating the trajectory offset coefficient of the ore particles based on the spatial contour features and the surface scattering features specifically includes:

[0018] The equivalent volume and centroid coordinates of the ore particles are calculated based on the aforementioned spatial contour features.

[0019] The compositional properties of ore particles are classified based on the surface scattering characteristics.

[0020] The equivalent volume, the centroid coordinates, and the composition attributes are input into a preset trajectory offset model, and the initial trajectory offset coefficient of the ore particles is output.

[0021] The initial trajectory offset coefficient is dynamically corrected based on the initial velocity of the ore particles to obtain the trajectory offset coefficient of the ore particles.

[0022] In some embodiments, predicting the trajectory partitioning of ore particles by using the trajectory offset coefficient and the local flow field parameters of the original ore flow specifically includes:

[0023] Local flow field parameters of the raw ore flow within the sorting channel are collected, including airflow velocity vector and pressure distribution.

[0024] The trajectory offset coefficient and the local flow field parameters are input into the particle motion dynamics model to calculate the motion acceleration of the ore particles under the action of airflow.

[0025] Based on the aforementioned motion acceleration and the initial motion state of the ore particles, the complete motion trajectory of the ore particles is simulated and generated.

[0026] The trajectory of the ore particles is divided into landing point zones based on the endpoint coordinates of the complete motion trajectory.

[0027] In some embodiments, mapping the landing point trajectory partition to an ore sorting window specifically includes:

[0028] Establish a spatial mapping relationship between the landing point trajectory partition and the physical coordinates of the sorting channel;

[0029] Based on the spatial mapping relationship, determine the intervention plane area corresponding to the landing trajectory partition within the sorting channel;

[0030] Based on the geometric boundary parameters of the intervention plane region, an ore sorting window is generated corresponding to the landing point trajectory partition.

[0031] In some embodiments, generating airflow intervention commands based on the decision-making influence relationship of the trajectory offset coefficient on the landing point trajectory partition within the ore sorting window specifically includes:

[0032] Establish the decision influence weight matrix of the trajectory offset coefficient and the landing point trajectory partition;

[0033] The airflow intervention intensity level corresponding to the ore sorting trajectory offset coefficient is determined based on the decision influence weight matrix.

[0034] The spatiotemporal parameter combination of the airflow pulse is generated based on the airflow intervention intensity level and the spatial position parameters of the ore sorting window;

[0035] The spatiotemporal parameter combination is converted into a driving signal for a multi-level pulsed air curtain, forming an airflow intervention command.

[0036] In some embodiments, determining the ore rejection airflow zone and the ore retention airflow zone in response to the airflow intervention command specifically includes:

[0037] Analyze the pressure gradient distribution in the airflow intervention command;

[0038] The threshold boundary for the intensity of airflow disturbance within the sorting channel is defined based on the pressure gradient distribution.

[0039] The region in the sorting channel where the pressure gradient value is greater than the threshold boundary is defined as the ore rejection airflow zone;

[0040] The region in the sorting channel where the pressure gradient value is less than or equal to the threshold boundary is defined as the ore retention gas flow zone.

[0041] Secondly, this application provides an intelligent ore pre-selection system based on airflow separation for performing an intelligent ore pre-selection method based on airflow separation. The system includes:

[0042] The feeding module is used to uniformly feed the raw ore stream into the sorting channel through a vibrating feeder, so that the raw ore stream forms a controllable sparse distribution in the free fall area.

[0043] The processing module is used to identify the spatial contour features and surface scattering features of ore particles upstream of the free fall zone, and generate the trajectory offset coefficient of the ore particles based on the spatial contour features and the surface scattering features.

[0044] The processing module is also used to predict the landing trajectory partitions of ore particles by using the trajectory offset coefficient and the local flow field parameters of the original ore flow, and to map the landing trajectory partitions as ore sorting windows. Within the ore sorting window, airflow intervention commands are generated based on the decision influence relationship of the trajectory offset coefficient on the landing trajectory partitions.

[0045] The processing module is also used to respond to the airflow intervention command, determine the ore rejection airflow zone and the ore retention airflow zone in the ore particle sorting process, and use a multi-stage pulse air curtain to send ore particles that meet the waste rock characteristics into the ore rejection airflow zone, while ore particles that do not meet the waste rock characteristics into the ore retention airflow zone.

[0046] The execution module is used to acquire the landing point response signal of ore particles through the receiving end set in the ore rejection airflow zone and the ore retention airflow zone after the airflow intervention, and update the airflow intervention command for subsequent ore particle sorting with the landing point response signal.

[0047] Thirdly, this application provides a computer device, the computer device including a memory and a processor, the memory storing code, the processor being configured to acquire the code and execute the above-described intelligent pre-selection method for ore based on airflow separation.

[0048] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described intelligent pre-selection method for ore based on airflow separation.

[0049] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects:

[0050] The intelligent pre-selection system and method for ore based on airflow separation provided in this application firstly involves feeding the raw ore stream uniformly into the separation channel using a vibrating feeder, causing the raw ore stream to form a controllable sparse distribution within the free fall zone. Secondly, the spatial contour features and surface scattering features of the ore particles are identified upstream of the free fall zone, and a trajectory offset coefficient for the ore particles is generated based on the spatial contour features and the surface scattering features. Further, the trajectory offset coefficient and the local flow field parameters of the raw ore stream are used to predict the landing point trajectory partitions of the ore particles, and these landing point trajectory partitions are mapped as ore separation windows. Within the ore separation windows... Based on the trajectory offset coefficient and its influence on the decision-making relationship of the landing point trajectory partition, an airflow intervention command is generated. Then, in response to the airflow intervention command, the ore rejection airflow zone and the ore retention airflow zone in the ore particle sorting process are determined. A multi-stage pulsed air curtain directs ore particles that meet the waste rock characteristics into the ore rejection airflow zone, while ore particles that do not meet the waste rock characteristics enter the ore retention airflow zone. Finally, after the airflow intervention, the landing point response signal of the ore particles is obtained by the receiving end set in the ore rejection airflow zone and the ore retention airflow zone, and the landing point response signal is used to update the airflow intervention command for subsequent ore particle sorting.

[0051] Therefore, this application can effectively sort ore particles under the interference of collisions between them. First, the raw ore flow is uniformly fed into the sorting channel by a vibrating feeder, forming a controllable sparse distribution in the free fall zone. This allows for precise control of the spacing and falling posture of the ore particles, effectively avoiding feature recognition interference caused by collisions and adhesion between particles, and providing a stable foundation for subsequent feature extraction and trajectory analysis. Second, the spatial contour features and surface scattering features of the ore particles are identified upstream of the free fall zone, and a trajectory offset coefficient is generated accordingly. This trajectory offset coefficient can transform the differences in the physical and optical properties of the particles into quantifiable numerical indicators, which can not only accurately distinguish the essential characteristics of waste rock and target ore, but also provide core parameter support for subsequent prediction of the particle motion trajectory under the action of airflow. Furthermore, the trajectory offset coefficient and the local flow field parameters of the raw ore flow are used to predict the landing trajectory partitions of the ore particles, which are then mapped to ore sorting windows. Within the window, airflow intervention commands are generated based on the decision-making influence relationship of the trajectory offset coefficient on the landing trajectory partitions, ensuring that the airflow energy is concentrated on the target particles and avoiding collisions between ore particles. The mechanical collisions between particles and the obstruction and disturbance of subsequent airflow by particle groups are addressed. Then, in response to airflow intervention commands, ore rejection airflow zones and ore retention airflow zones are determined. Multi-stage pulsed air curtains push ore particles that meet the waste rock characteristics into the rejection airflow zone, while allowing ore particles that do not meet the waste rock characteristics to smoothly enter the retention airflow zone. This zoning process allows for differentiated intervention of ore particles with different properties, ensuring efficient separation and rejection of waste rock while avoiding unnecessary airflow disturbance to the target ore. This results in a large number of ore particles being effectively sorted into their corresponding zones, thereby improving the separation of waste rock from the target ore. The separation efficiency and purity of the target ore are improved. Finally, after airflow intervention, the landing point response signal of the ore particles is obtained by the receiving end set in the ore rejection airflow zone and the ore retention airflow zone, and the subsequent airflow intervention command is updated accordingly. This allows the system to flexibly adapt to complex working conditions such as fluctuations in ore composition and changes in equipment operating status, continuously optimize the sorting accuracy, and ultimately achieve intelligent control of the entire ore pre-selection process, greatly improving the efficiency and reliability of pre-selection operations. In summary, the technical solution provided in this application can effectively sort ore particles under collision interference between ore particles. Attached Figure Description

[0052] Figure 1 This is an exemplary flowchart of an intelligent pre-selection method for ores based on airflow separation, according to some embodiments of this application;

[0053] Figure 2 This is an exemplary flowchart illustrating the determination of trajectory offset coefficients according to some embodiments of this application;

[0054] Figure 3 This is a schematic diagram of the structure of an intelligent ore pre-selection system based on airflow separation, according to some embodiments of this application;

[0055] Figure 4 This is a schematic diagram of the structure of a computer device for implementing an intelligent pre-selection method for ore based on airflow separation, according to some embodiments of this application. Detailed Implementation

[0056] To better understand the technical solution of this application, the technical solution of this application will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0057] refer to Figure 1 The figure is an exemplary flowchart of an intelligent pre-selection method for ore based on airflow separation according to some embodiments of this application. The figure mainly includes the following steps:

[0058] In step S101, the raw ore stream is uniformly fed into the sorting channel by a vibrating feeder, so that the raw ore stream forms a controllable sparse distribution in the free fall area.

[0059] In practice, firstly, the vibrating feeder is started, and the vibrating feeder periodically guides the raw ore flow evenly to the sorting channel. In the initial stage when the raw ore flow begins to fall freely, a uniform airflow field is generated by the airflow buffer device (such as a low-pressure stabilizing nozzle) in the sorting channel to buffer the raw ore flow, offsetting the adsorption and friction between ore particles, so that the raw ore flow forms a controllable sparse distribution without obvious agglomeration in the free fall area.

[0060] It should be noted that, in this application, the raw ore stream refers to an aggregate of ore particles obtained from the mining process that has not undergone subsequent purification or sorting. The vibrating feeder in this application is the core equipment for achieving uniform conveying and initial control of the raw ore stream. Essentially, it is a special device that uses the principle of mechanical vibration to complete the quantitative and directional conveying of materials. Structurally, it typically includes four parts: a vibrating motor (or vibrator), a feeding trough, a shock-absorbing device, and a control system. The vibrating motor provides periodic vibration power to the equipment. The feeding trough adopts an inclined design (the inclination angle can be adjusted according to the characteristics of the ore) to guide the ore flow. The shock-absorbing device can reduce the impact of vibration on the equipment base and surrounding structures. The control system can adjust parameters such as vibration frequency and amplitude in real time.

[0061] In step S102, the spatial contour features and surface scattering features of the ore particles are identified upstream of the free fall zone, and the trajectory offset coefficient of the ore particles is generated based on the spatial contour features and the surface scattering features.

[0062] In some embodiments, identifying the spatial profile and surface scattering characteristics of ore particles upstream of the free fall zone is achieved using the following steps:

[0063] Multi-view synchronous imaging of ore particles was performed upstream of the free fall zone using a high-speed linear array camera to obtain the original image dataset of the ore particles.

[0064] The original image dataset is preprocessed to obtain effective images of ore particles;

[0065] An edge detection algorithm is used to extract the spatial contour features of ore particles from the valid image;

[0066] Spectral analysis is performed on the effective image to obtain the scattered light intensity distribution data on the surface of the ore particles;

[0067] The surface scattering characteristics of the ore particles are determined based on the scattered light intensity distribution data.

[0068] In specific implementation, firstly, a high-speed linear array camera performs multi-view synchronous imaging of ore particles upstream of the free fall zone to obtain an original image dataset. This original image dataset refers to an initial set of images containing ore particles and background information, including ore particle images from different perspectives. Secondly, for each ore particle image in the original image dataset, an adaptive threshold segmentation method is first used to separate the ore particles from the background. Then, morphological filtering (such as erosion and dilation operations) is used to remove noise and edge burrs, extracting the foreground region of the ore particles to obtain an effective image. This effective image refers to an image containing clear outlines and surface information of the ore particles. Further, the Canny edge detection algorithm in image processing is used to first calculate the image gradient of the effective image to identify edge intensity and direction. Then, non-maximum suppression is used to refine the edges, and a double threshold method is used to filter the edges. The true edges are selected, and then the spatial contour features of the ore particles are extracted. The spatial contour features refer to the set of geometric parameters describing the boundary of the ore particle shape, and the spatial contour features refer to the continuous lines reflecting the boundary of the ore particle shape. Then, the radiation intensity value of each pixel in the foreground region of the ore particle in the effective image is extracted using existing spectral imaging technology to obtain the scattered light intensity distribution data of the ore particle surface. The scattered light intensity distribution data refers to the spatial distribution information of the reflected light intensity of the ore particle surface in the spectral band. Finally, the surface scattering characteristics of the ore particles are determined based on the scattered light intensity distribution data. That is, the proportion of different radiation intensity values ​​in the scattered light intensity distribution data is used as the input variable to the information entropy function, and the output result of the information entropy function is used as the surface scattering characteristics of the ore particles. The surface scattering characteristics refer to the parameters reflecting the disordered changes of the material on the surface of the ore particles.

[0069] It should be noted that by determining the spatial contour features and surface scattering features of ore particles, key physical and optical properties can be provided for intelligent ore pre-selection. Spatial contour features can accurately reflect the geometric shape of ore particles (such as particle size, shape regularity, surface area, etc.), while surface scattering features contain information about the ore composition and can serve as an important indicator to distinguish useful minerals from waste rock. The feature dataset formed by the combination of the two can not only provide the particle's own physical parameters to support the subsequent generation of trajectory offset coefficients, but also provide a classification basis for airflow intervention decisions through compositional attribute judgment, thereby achieving accurate identification and sorting guidance of ore particles and ensuring the effective separation of target minerals and waste rock during the pre-selection process.

[0070] In some embodiments, reference Figure 2 As shown, this figure is an exemplary flowchart of determining the trajectory offset coefficient according to some embodiments of this application. In this embodiment, the trajectory offset coefficient of the ore particles can be generated based on the spatial contour features and the surface scattering features by the following steps:

[0071] In step S1021, the equivalent volume and centroid coordinates of the ore particles are calculated based on the spatial contour features.

[0072] In step S1022, the compositional properties of the ore particles are classified based on the surface scattering characteristics.

[0073] In step S1023, the equivalent volume, the centroid coordinates, and the composition attribute classification are input into a preset trajectory offset model, and the initial trajectory offset coefficient of the ore particles is output.

[0074] In step S1024, the initial trajectory offset coefficient is dynamically corrected based on the initial velocity of the ore particles to obtain the trajectory offset coefficient of the ore particles.

[0075] In specific implementation, firstly, a three-dimensional mesh model of the particles is constructed using the boundary point coordinates in the spatial contour features. Then, the Monte Carlo integration method is used to calculate the spatial volume enclosed by the three-dimensional mesh model as the equivalent volume. The equivalent volume refers to the volume of an ideal geometric body equal to the actual volume of the ore particles. Simultaneously, the centroid coordinates are obtained by calculating the weighted average of the coordinates of all vertices in the three-dimensional mesh model. The centroid coordinates are the three-dimensional spatial coordinates representing the position of the mass center of the ore particles. Secondly, when determining the compositional attribute classification of ore particles based on surface scattering characteristics, the surface scattering characteristics are compared with a standard spectral feature library of known mineral components. The correlation coefficient analysis method is used to calculate the similarity between the surface scattering characteristics and the standard spectral features in the standard spectral feature library of known mineral components. The mineral component corresponding to the standard spectral feature with the highest similarity is determined as the compositional attribute classification of the ore particles. The compositional attribute classification refers to the category divided according to the types of minerals contained in the ore particles. Then, when the equivalent volume, centroid coordinates, and compositional attribute classification are input into a preset trajectory offset model, which is constructed based on particle dynamics theory, the equivalent volume is first converted into particle mass (obtained by multiplying by the known mineral density). Then, combining the centroid coordinates, the point of application of gravity on the ore particles is calculated. Simultaneously, based on the compositional properties, a preset air resistance coefficient is invoked, and the basic coefficient of particle displacement in the airflow is derived using Newton's second law. The initial trajectory displacement coefficient of the ore particles is output, which refers to the calculated trajectory displacement value without considering the influence of the initial velocity. Finally, when dynamically correcting the initial trajectory displacement coefficient based on the initial velocity of the ore particles, a high-speed camera system captures the velocity vector of the particles entering the sorting channel. This velocity vector is decomposed into horizontal and vertical components and represented... The additional influence of the initial velocity on the trajectory deviation is calculated by incorporating it into the kinematic equations. The trajectory deviation coefficient of the ore particle is obtained by adding the additional influence to the initial trajectory deviation coefficient. Specifically, the kinematic equations refer to the classical mechanics equations that describe the relationship between the position of an object in space and its velocity and acceleration over time. The additional influence is calculated by decomposing the velocity vector into horizontal and vertical components and substituting them into the kinematic equations. The additional influence is the resultant displacement of the horizontal additional displacement generated by the horizontal component within the airflow intervention time and the vertical additional displacement generated by the vertical component within the same time.

[0076] It should be noted that the trajectory deviation coefficient in this application refers to a quantitative parameter of the degree to which ore particles deviate from their natural trajectory under the action of airflow. Existing technologies mostly calculate trajectory deviation parameters solely based on particle size or density, ignoring the implicit influence of compositional differences on aerodynamic characteristics. In contrast, this scheme constructs a multi-dimensional parameter-coupled calculation model based on the equivalent volume and centroid coordinates obtained from spatial contour features, combined with the compositional attribute classification determined by surface scattering features. This model incorporates physical parameters such as particle inertia and force equilibrium point, as well as a composition-related air resistance coefficient correction term, significantly improving the adaptability of the coefficient to different types of ore particles. It breaks through the limitations of existing technologies dominated by a single parameter, forming a more accurate trajectory deviation quantification standard. This provides a more reliable decision-making basis for the generation of subsequent airflow intervention commands, demonstrating significant technological progress in both sorting accuracy and adaptability.

[0077] In step S103, the trajectory partitions of ore particles are predicted by the trajectory offset coefficient and the local flow field parameters of the original ore flow, and the trajectory partitions are mapped as ore sorting windows. Within the ore sorting windows, airflow intervention commands are generated based on the decision influence relationship of the trajectory offset coefficient on the trajectory partitions.

[0078] In some embodiments, the prediction of the trajectory partitioning of ore particles by means of the trajectory offset coefficient and the local flow field parameters of the original ore flow is achieved by the following steps:

[0079] Local flow field parameters of the raw ore flow within the sorting channel are collected, including airflow velocity vector and pressure distribution.

[0080] The trajectory offset coefficient and the local flow field parameters are input into the particle motion dynamics model to calculate the motion acceleration of the ore particles under the action of airflow.

[0081] Based on the aforementioned motion acceleration and the initial motion state of the ore particles, the complete motion trajectory of the ore particles is simulated and generated.

[0082] The trajectory of the ore particles is divided into landing point zones based on the endpoint coordinates of the complete motion trajectory.

[0083] In specific implementation, firstly, when collecting local flow field parameters of the raw ore flow within the sorting channel, a miniature hot-wire anemometer and pressure sensor array arranged on the inner wall of the sorting channel are used to collect airflow velocity vectors (including velocity magnitude and direction) and pressure distributions at different locations in real time, and combine them into local flow field parameters. These local flow field parameters refer to the airflow motion state and pressure change information within the sorting channel. Secondly, when inputting the trajectory offset coefficient and the local flow field parameters into a particle motion dynamics model to calculate the motion acceleration, this model is constructed based on the Euler-Granger method. First, the trajectory offset coefficient is converted into a particle response coefficient to the airflow. Then, the aerodynamic force on the particle is calculated by combining the airflow velocity vector in the local flow field parameters. Based on this aerodynamic force, the motion acceleration is derived using Newton's second law. The particle motion dynamics model is a mathematical model describing the motion changes of particles under mechanical action in the airflow. The motion acceleration... The degree refers to the change in particle velocity per unit time. Then, when simulating and generating a complete motion trajectory based on the motion acceleration and the initial motion state of the ore particles, the initial motion state includes the initial position of the particles when they enter the sorting channel. By substituting the motion acceleration into the kinematic integral equation (i.e., integrating velocity with time to obtain the displacement equation), the position coordinates of the ore particles in three-dimensional space are calculated time by time. The position coordinates at each time point are connected from the initial position to form a complete motion trajectory. The initial motion state refers to the spatial position information of the particles when they begin to participate in sorting. The complete motion trajectory refers to the entire motion path of the ore particles from entering the sorting channel to reaching the discharge area. Finally, when dividing the discharge trajectory into partitions according to the endpoint coordinates of the complete motion trajectory, a grid coordinate system of the discharge area is preset. The endpoint coordinates of each particle trajectory are matched to the corresponding grid cell. The discharge area is divided into multiple regions according to the spatial distribution of the grid cells. Each region is a discharge trajectory partition.

[0084] It should be noted that, in this application, the landing trajectory zoning refers to the spatial area divided according to the final landing position of the ore particles. The purpose of determining the landing trajectory zoning is to provide a precise spatial positioning basis for airflow intervention in ore sorting. By clarifying the expected landing range of different ore particles in the landing area, subsequent airflow intervention operations can be precisely applied to specific areas, avoiding sorting errors caused by excessive or insufficient intervention range. At the same time, the landing trajectory zoning associates the movement trajectory of ore particles with the spatial area, providing a direct reference for dividing the ore rejection airflow zone and the ore retention airflow zone, ensuring that ore particles that meet the characteristics of waste rock and non-waste rock can be accurately guided to the corresponding areas.

[0085] In some embodiments, mapping the landing point trajectory partition to an ore sorting window can be achieved using the following steps:

[0086] Establish a spatial mapping relationship between the landing point trajectory partition and the physical coordinates of the sorting channel;

[0087] Based on the spatial mapping relationship, determine the intervention plane area corresponding to the landing trajectory partition within the sorting channel;

[0088] Based on the geometric boundary parameters of the intervention plane region, an ore sorting window is generated corresponding to the landing point trajectory partition.

[0089] In specific implementation, firstly, when establishing the spatial mapping relationship between the landing trajectory partition and the physical coordinates of the sorting channel, a three-dimensional rectangular coordinate system is constructed with the initial point at the entrance of the sorting channel as the origin. The grid coordinates of the landing trajectory partition are then converted into physical coordinates within the sorting channel using a coordinate transformation matrix (homogeneous coordinate transformation can be used). This establishes the spatial mapping relationship between the landing trajectory partition and the physical coordinates of the sorting channel. This spatial mapping relationship refers to a mathematical model characterizing the correspondence between the actual spatial positions of the landing trajectory partition and the sorting channel. Secondly, when determining the intervention plane area corresponding to the landing trajectory partition within the sorting channel based on the spatial mapping relationship, the complete motion trajectory of the ore particles within the sorting channel is considered. The highest influence section (i.e., the plane where the ore particle movement direction intersects with the preset plane, the preset plane refers to several parallel reference planes that are pre-set in the sorting channel design stage and are perpendicular to the mainstream particle movement direction) is then projected onto the highest influence section through spatial mapping relationship to form a closed polygonal region (i.e., the intervention plane region). The intervention plane region refers to the planar spatial range within the sorting channel where effective intervention can be applied to the particle movement trajectory. Finally, the vertex coordinates of the boundary of the intervention plane region are extracted, the boundary contour is smoothed by fitting a B-spline curve, and the fitted closed region is used as the ore sorting window corresponding to the drop point trajectory region.

[0090] It should be noted that, in this application, the ore sorting window refers to the spatial area within the sorting channel where airflow intervention can be implemented to change the particle trajectory. The determination of the ore sorting window provides a precise spatial range for airflow intervention. By mapping the landing trajectory to specific planar areas within the sorting channel, the effective spatial boundary for airflow intervention is clarified, ensuring that the effect of the multi-stage pulse air curtain can be concentrated on the critical path of the target particle's trajectory, thus avoiding ineffective energy loss of airflow.

[0091] In some embodiments, generating airflow intervention commands based on the decision-making influence relationship of the trajectory offset coefficient on the landing point trajectory partition within the ore sorting window can be achieved through the following steps:

[0092] Establish the decision influence weight matrix of the trajectory offset coefficient and the landing point trajectory partition;

[0093] The airflow intervention intensity level corresponding to the ore sorting trajectory offset coefficient is determined based on the decision influence weight matrix.

[0094] The spatiotemporal parameter combination of the airflow pulse is generated based on the airflow intervention intensity level and the spatial position parameters of the ore sorting window;

[0095] The spatiotemporal parameter combination is converted into a driving signal for a multi-level pulsed air curtain, forming an airflow intervention command.

[0096] In specific implementation, firstly, the analytic hierarchy process (AHP) is used to treat the trajectory offset coefficient and the trajectory partitions of each landing point as elements of a hierarchical structure. Existing scaling methods are used to compare elements pairwise, and the comparison results are normalized and used as decision influence weight values. All decision influence weight values ​​are arranged by rows (trajectory offset coefficients) and columns (trajectory distribution areas) to form a decision influence weight matrix. This matrix reflects the degree of influence of different trajectory offset coefficients on the decision-making of each landing point trajectory partition. Secondly, the trajectory offset coefficient of the current ore particle is used as a row vector, and matrix multiplication is performed with the decision influence weight matrix. The result is used as the correlation value of the landing point trajectory partition. A preset correlation-intensity level mapping table (obtained through fitting previous experimental data, not detailed here) is called to identify the airflow intervention intensity level corresponding to the correlation value. The airflow intervention intensity level refers to the magnitude of the change in particle trajectory caused by airflow. Then, when generating the spatiotemporal parameter combination of airflow pulses based on the airflow intervention intensity level and the spatial position parameters of the ore sorting window, the... The spatial location parameters include the lateral width and geometric center of the ore sorting window. A reference pressure value is retrieved from a preset pressure database based on the airflow intervention intensity level. The pulse duration is calculated by combining the length of the ore sorting window along the particle movement direction with the real-time particle speed (acquired via a high-speed camera system). The lateral width of the ore sorting window is used as the nozzle arrangement range, which refers to the spatial distribution interval of the nozzles participating in the airflow intervention. The reference pressure value, pulse duration, and nozzle arrangement range are combined to form a spatiotemporal parameter combination for the airflow pulse. This spatiotemporal parameter combination refers to the set of pressure parameters, action time parameters, and spatial action range parameters describing the airflow pulse. Finally, a digital-to-analog converter converts the reference pressure value parameter in the spatiotemporal parameter combination into a corresponding current signal, the pulse duration parameter into a timer trigger signal, and the nozzle arrangement range parameter into a digital signal. The current signal, trigger signal, and digital signal are then combined to form an airflow intervention command.

[0097] It should be noted that the airflow intervention command in this application refers to the set of electrical signals used to drive the multi-stage pulse air curtain to work according to the set parameters. The determination of the airflow intervention command is the core link connecting the ore particle characteristic analysis and the actual sorting operation. Its function is to transform abstract parameters such as trajectory offset coefficient, landing point trajectory partition and ore sorting window into specific executable operation signals. By specifying the pressure, duration and nozzle arrangement range of the airflow pulse, the command ensures that the multi-stage pulse air curtain can accurately apply directional force within the ore sorting window, so that the ore particles to be removed deviate from the original trajectory as expected and enter the corresponding landing point partition, while the retained ore particles are not disturbed. This not only realizes the differentiated processing of ore particles with different properties, but also ensures the stability and consistency of airflow intervention through standardized driving signals, providing a direct operational basis for improving the accuracy, efficiency and automation level of ore sorting.

[0098] In step S104, in response to the airflow intervention command, the ore rejection airflow zone and the ore retention airflow zone in the ore particle sorting process are determined, and ore particles that meet the waste rock characteristics are sent into the ore rejection airflow zone by a multi-stage pulse air curtain, while ore particles that do not meet the waste rock characteristics are sent into the ore retention airflow zone.

[0099] In some embodiments, determining the ore rejection airflow zone and the ore retention airflow zone in response to the airflow intervention command during the ore particle sorting process can be achieved by the following steps:

[0100] Analyze the pressure gradient distribution in the airflow intervention command;

[0101] The threshold boundary for the intensity of airflow disturbance within the sorting channel is defined based on the pressure gradient distribution.

[0102] The region in the sorting channel where the pressure gradient value is greater than the threshold boundary is defined as the ore rejection airflow zone;

[0103] The region in the sorting channel where the pressure gradient value is less than or equal to the threshold boundary is defined as the ore retention gas flow zone.

[0104] In specific implementation, firstly, the output pressure values ​​of each nozzle are extracted from the pressure parameters included in the airflow intervention command. Combined with the spatial coordinate distribution of the nozzles within the sorting channel, a three-dimensional pressure field model within the channel is constructed using inverse distance weighted interpolation. The pressure gradient distribution is obtained by calculating the spatial derivative of the pressure field model (calculating the ratio of the pressure difference between adjacent points to the distance). This pressure gradient distribution describes the spatial rate of change of pressure within the sorting channel. Secondly, the pressure gradient values ​​are used as a quantitative indicator of the airflow disturbance intensity. An adaptive threshold calculation is performed on the pressure gradient distribution data using the Otsu threshold segmentation algorithm. A threshold boundary for the intensity of airflow disturbance within the sorting channel is obtained. This threshold boundary refers to the dividing parameter that defines the strength of airflow disturbance within the sorting channel. Then, the region in the sorting channel with a pressure gradient value greater than the threshold boundary is designated as the ore rejection airflow zone. This ore rejection airflow zone refers to the airflow zone that can exert sufficient force on ore particles to deviate from their original trajectory to the rejection area. Finally, the region in the sorting channel with a pressure gradient value less than or equal to the threshold boundary is designated as the ore retention airflow zone. This ore retention airflow zone refers to the airflow zone where airflow disturbance will not significantly change the original trajectory of the ore particles.

[0105] It should be noted that this application provides a clear standard for dividing the airflow action area in the ore sorting process by defining the ore rejection airflow zone and the ore retention airflow zone. The ore rejection airflow zone, with its airflow disturbance intensity exceeding the threshold, can apply sufficient force to the ore particles to be rejected, ensuring that they deviate from their original trajectory and enter the waste rock collection area as expected. The ore retention airflow zone, because its disturbance intensity does not exceed the standard, can avoid significantly interfering with the movement of the target ore particles, ensuring that they smoothly enter the useful mineral collection area. This zoning method not only achieves precise diversion of ore particles with different properties, but also reduces energy waste through clear airflow action boundaries, and provides a clear regional reference for the subsequent evaluation of sorting effect, thereby improving the accuracy and efficiency of ore sorting.

[0106] In some embodiments, the process of using a multi-stage pulsed air curtain to direct ore particles that meet the waste rock characteristics into the ore removal airflow zone, while allowing ore particles that do not meet the waste rock characteristics to enter the ore retention airflow zone, can be achieved through the following steps:

[0107] Mark the target particles that meet the characteristics of waste rock;

[0108] A multi-stage pulsed air curtain is driven to apply a directional force to target ore particles that meet the characteristics of waste rock. The target particles, affected by the pulsed airflow, change their trajectory and enter the ore removal airflow zone.

[0109] Unmarked ore particles that do not meet the characteristics of waste rock enter the ore retention gas flow zone along their original trajectory.

[0110] It should be noted that the multi-stage pulse air curtain in this application consists of multiple independently controlled solenoid valves and nozzles, and each nozzle can independently adjust the air pressure and pulse duration.

[0111] In practice, firstly, real-time images of ore particles are acquired using a high-speed industrial camera. Combined with previously determined surface scattering and spatial contour features, a template matching algorithm is used to compare the particle features (i.e., surface scattering and spatial contour features) with preset waste rock feature templates. The normalized cross-correlation coefficient is calculated as the waste rock matching degree, and ore particles with a matching degree greater than the preset matching degree are selected as target particles that meet the waste rock characteristics. These target particles, which meet the waste rock characteristics, refer to ore particles possessing preset physical and optical properties of waste rock. Secondly, when a multi-stage pulsed air curtain is driven to apply directional force to the target ore particles that meet the waste rock characteristics, the corresponding air... The flow intervention command (including pressure, pulse duration, and nozzle combination parameters) controls a multi-stage pulsed air curtain via a solenoid valve assembly. The impact force generated by the air jet applies a directional force (such as a lateral force deviating from the ore retention zone) to the target particles. The target particles, affected by the pulsed airflow, change their trajectory and enter the ore rejection airflow zone. The directional force refers to the force that can change the direction of particle movement. Finally, when unmarked ore particles that do not meet the waste rock characteristics enter the ore retention airflow zone along their original trajectory, these particles, because they have not triggered the air curtain drive signal, are not subject to additional airflow forces. They maintain their initial motion state only under the influence of gravity and the basic airflow (stable airflow that has not been disturbed by the pulse) within the sorting channel, and enter the ore retention airflow zone along their original trajectory.

[0112] In step S105, after the airflow intervention, the landing point response signal of the ore particles is obtained by the receiving end set in the ore rejection airflow zone and the ore retention airflow zone, and the airflow intervention command for subsequent ore particle sorting is updated with the landing point response signal.

[0113] In some embodiments, after the airflow intervention, the landing point response signal of the ore particles can be obtained by the following steps through the receiving end set in the ore removal airflow zone and the ore retention airflow zone:

[0114] An array of sensor units is deployed at the material drop path receiving ends of the ore rejection airflow zone and the ore retention airflow zone, respectively.

[0115] After the airflow intervention, the array-type sensing unit captures in real time the impact signal and obstruction signal generated by the ore particles falling into the ore removal airflow zone and the ore retention airflow zone;

[0116] The impact signal and the blocking signal are converted from analog to digital to generate the impact point response signal of the ore particles.

[0117] In specific implementation, firstly, when deploying array-type sensing units at the receiving ends of the material drop paths in the ore rejection airflow zone and the ore retention airflow zone, an array-type sensing unit composed of piezoelectric sensors and infrared pairs is selected. The piezoelectric sensors are used to detect particle impact, and the infrared pairs are used to detect particle obstruction. The array-type sensing unit refers to a detection device formed by arranging multiple sensors of the same type. Then, after the airflow intervention, the piezoelectric sensors and infrared pairs in the array-type sensing unit capture in real time the impact signals and obstruction signals generated by ore particles falling into the ore rejection airflow zone and the ore retention airflow zone. The impact signal refers to the electrical signal generated by the particle hitting the sensor, and the obstruction signal refers to the electrical signal generated by the particle blocking the optical path. Finally, the impact signal and obstruction signal can be converted from analog to digital by an analog-to-digital converter, and the converted signals are combined into the ore particle drop point response signal.

[0118] It should be noted that, in this application, the landing point response signal refers to the set of electrical signals reflecting the basic information of the landing point of ore particles after they fall into the material dropping area. The landing point response signal includes electrical signals after analog-to-digital conversion of impact signals and obstruction signals. The determination of the landing point response signal provides a key basis for real-time monitoring of ore sorting effect and subsequent parameter optimization. As an instantaneous feedback of particle landing point by the array-type sensing unit, it includes the original electrical signals after analog-to-digital conversion of impact signals and obstruction signals. It can intuitively reflect the actual distribution status of ore particles that meet the characteristics of waste rock and non-waste rock in the rejection area and retention area. It can not only be used to verify the execution accuracy of airflow intervention commands, but also provide original data support for correcting trajectory offset coefficients and adjusting airflow intervention intensity levels, thereby improving the stability and adaptability of the ore sorting system.

[0119] In some embodiments, updating the airflow intervention command for subsequent ore particle sorting with the landing point response signal can be achieved by the following steps:

[0120] The actual separation ratio of ore particles in the ore rejection gas flow zone and the ore retention gas flow zone is obtained based on the landing point response signal.

[0121] The deviation between the actual sorting ratio and the preset sorting threshold is calculated to obtain the sorting accuracy error value;

[0122] The intervention command adjustment coefficient during ore particle sorting is determined based on the sorting accuracy error value.

[0123] The mapping relationship between the original trajectory offset coefficient and the local flow field parameters is corrected by the intervention command adjustment coefficient, and the airflow intervention command for subsequent ore particle sorting is regenerated based on the corrected mapping relationship.

[0124] In specific implementation, firstly, the number of particles falling into the ore rejection airflow zone and the ore retention airflow zone is extracted from the landing point response signal (achieved by counting the number of sensor triggers in different zones). A division operation is used to calculate the ratio of the number of particles in the ore rejection airflow zone to the total number of particles, and the ratio of the number of particles in the ore retention airflow zone to the total number of particles, to obtain the actual sorting ratio of ore particles in the ore rejection airflow zone and the ore retention airflow zone. The actual sorting ratio refers to the proportion of the number of particles actually falling into the rejection zone and the retention zone to the total number of particles. Secondly, a subtraction operation is used to calculate the difference between the actual sorting ratio and the preset sorting threshold. The absolute value of the difference is then taken to obtain the sorting accuracy error value. The preset sorting threshold is a target ratio value set according to the sorting requirements, which can be set according to actual needs and is not limited here. The sorting accuracy error value refers to the quantified value of the deviation between the actual sorting ratio and the preset target ratio. Then, a proportional-integral control algorithm is used to adjust the sorting accuracy. The error value is used as input. The results calculated by the proportional stage (outputting adjustment amount proportionally to the error size) and the integral stage (accumulating error to eliminate static deviation) are used as the intervention command adjustment coefficient during ore particle sorting. The intervention command adjustment coefficient refers to the proportional coefficient used to correct the original airflow intervention command parameters. Finally, the mapping relationship is a pre-established table of relationship between trajectory offset coefficients and local flow field parameters. The intervention command adjustment coefficients are coupled with the parameters in this mapping relationship through multiplication to obtain the corrected mapping relationship. The parameters specifically refer to the correlation weight values ​​used to associate the two in the mapping relationship table of the original trajectory offset coefficients and local flow field parameters. Then, the airflow intervention command for subsequent ore particle sorting is regenerated based on the corrected mapping relationship according to the original airflow intervention command generation logic. The mapping relationship between the original trajectory offset coefficients and local flow field parameters refers to the preset data relationship describing the corresponding association between the two. The table of relationship between the trajectory offset coefficients and local flow field parameters is as follows:

[0125] Trajectory offset coefficient\Local flow field parameters (0,0.2] (0.2,0.4] (0.4,0.6] (0.6,0.8] (0.8,1.0) (0,0.2] 0.1 0.2 0.3 0.4 0.5 (0.2,0.4] 0.3 0.4 0.5 0.6 0.7 (0.4,0.6] 0.5 0.6 0.7 0.8 0.9 (0.6,0.8] 0.7 0.8 0.9 1.0 1.0 (0.8,1.0) 0.9 1.0 1.0 1.0 1.0

[0126] In the above relationship table, the numerical ranges of the trajectory offset coefficient and the local flow field parameter are the normalized trajectory offset coefficient and the normalized local flow field parameter, respectively. The purpose of this relationship table is that the larger the trajectory offset coefficient and the higher the local flow field parameter, the larger the correlation weight value, indicating a stronger airflow intervention.

[0127] It should be noted that the corrected mapping relationship in this embodiment refers to the correspondence between the trajectory offset coefficient after being corrected by the intervention command adjustment coefficient and the local flow field parameters. By correcting the mapping relationship, the conversion ratio of the trajectory offset coefficient to the local flow field parameters can be adjusted in a targeted manner to avoid mapping deviations caused by factors such as equipment wear and ore property fluctuations. At the same time, the corrected mapping relationship serves as the benchmark for generating subsequent airflow intervention commands, ensuring that the new commands more accurately match the current particle sorting requirements and reducing the accumulation of sorting accuracy errors.

[0128] Furthermore, in another aspect of this application, in some embodiments, this application provides an intelligent ore pre-selection system based on airflow separation, referencing... Figure 3 The figure is a schematic diagram of the structure of an intelligent ore pre-selection system based on airflow separation according to some embodiments of this application. The intelligent ore pre-selection system based on airflow separation includes: a feeding module 201, a processing module 202, and an execution module 203, which are described below:

[0129] Feeding module 201, in this application, is mainly used to feed the raw ore stream into the sorting channel evenly through a vibrating feeder, so that the raw ore stream forms a controllable sparse distribution in the free fall area.

[0130] Processing module 202, in this application, is mainly used to identify the spatial contour features and surface scattering features of ore particles upstream of the free fall zone, and to generate the trajectory offset coefficient of ore particles based on the spatial contour features and the surface scattering features.

[0131] The processing module 202 is further configured to predict the landing trajectory partition of ore particles by means of the trajectory offset coefficient and the local flow field parameters of the original ore flow, and map the landing trajectory partition as an ore sorting window. Within the ore sorting window, an airflow intervention command is generated based on the decision influence relationship of the trajectory offset coefficient on the landing trajectory partition.

[0132] In addition, the processing module 202 is also used to respond to the airflow intervention command, determine the ore rejection airflow zone and the ore retention airflow zone in the ore particle sorting process, and use a multi-stage pulse air curtain to send ore particles that meet the waste rock characteristics into the ore rejection airflow zone, while ore particles that do not meet the waste rock characteristics into the ore retention airflow zone.

[0133] The execution module 203 in this application is mainly used to obtain the landing point response signal of ore particles by means of the receiving end set in the ore rejection airflow zone and the ore retention airflow zone after the airflow intervention, and update the airflow intervention command for subsequent ore particle sorting with the landing point response signal.

[0134] In addition, this application also provides a computer device, the computer device including a memory and a processor, the memory storing code, the processor being configured to acquire the code and execute the above-described intelligent pre-selection method for ore based on airflow separation.

[0135] In some embodiments, reference Figure 4 The figure is a schematic diagram of a computer device for implementing an intelligent ore pre-selection method based on airflow separation, according to some embodiments of this application. The intelligent ore pre-selection method based on airflow separation in the above embodiments can be implemented through... Figure 4 The computer device shown is used to implement this, and the computer device includes at least one processor 301, a communication bus 302, a memory 303, and at least one communication interface 304.

[0136] The processor 301 may be a general-purpose central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more devices used to control the execution of the intelligent pre-selection method for ore based on airflow separation in this application.

[0137] The communication bus 302 can be used to transmit information between the aforementioned components.

[0138] The memory 303 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disks or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. The memory 303 may exist independently and be connected to the processor 301 via the communication bus 302. The memory 303 may also be integrated with the processor 301.

[0139] The memory 303 stores program code for executing the scheme of this application, and its execution is controlled by the processor 301. The processor 301 executes the program code stored in the memory 303. The program code may include one or more software modules. In the above embodiments, the determination of the intelligent pre-selection method for ore based on airflow separation can be implemented by the processor 301 and one or more software modules in the program code in the memory 303.

[0140] Communication interface 304 uses any transceiver-like device for communicating with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), etc.

[0141] In a specific implementation, as one example, a computer device may include multiple processors, each of which may be a single-core (single-CPU) processor or a multi-core (multi-CPU) processor. Here, a processor may refer to one or more devices, circuits, and / or processing cores used to process data (e.g., computer program instructions).

[0142] The aforementioned computer device can be a general-purpose computer device or a special-purpose computer device. In specific implementations, the computer device can be a desktop computer, a portable computer, a network server, a handheld digital assistant (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device, or an embedded device. This application does not limit the type of computer device.

[0143] In addition, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described intelligent pre-selection method for ore based on airflow separation.

[0144] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0145] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A smart pre-selection method for ores based on airflow separation, characterized in that, Includes the following steps: The raw ore stream is fed evenly into the sorting channel by a vibrating feeder, so that the raw ore stream forms a controllable sparse distribution in the free fall area; The spatial contour features and surface scattering features of ore particles are identified upstream of the free fall zone, and the trajectory offset coefficient of the ore particles is generated based on the spatial contour features and the surface scattering features. The trajectory deviation coefficient and the local flow field parameters of the original ore flow are used to predict the landing trajectory partitions of ore particles, and the landing trajectory partitions are mapped as ore sorting windows. Within the ore sorting window, airflow intervention commands are generated based on the decision influence relationship of the trajectory deviation coefficient on the landing trajectory partitions. In response to the airflow intervention command, the ore rejection airflow zone and the ore retention airflow zone in the ore particle sorting process are determined, and the ore particles that meet the waste rock characteristics are entered into the ore rejection airflow zone by the multi-stage pulse air curtain, while the ore particles that do not meet the waste rock characteristics are entered into the ore retention airflow zone. After the airflow intervention, the landing point response signal of the ore particles is obtained by the receiving end set in the ore rejection airflow zone and the ore retention airflow zone, and the airflow intervention command for subsequent ore particle sorting is updated with the landing point response signal.

2. The method as described in claim 1, characterized in that, The identification of spatial contour features and surface scattering features of ore particles upstream of the free fall zone specifically includes: Multi-view synchronous imaging of ore particles was performed upstream of the free fall zone using a high-speed linear array camera to obtain the original image dataset of the ore particles. The original image dataset is preprocessed to obtain effective images of ore particles; An edge detection algorithm is used to extract the spatial contour features of ore particles from the valid image; Spectral analysis is performed on the effective image to obtain the scattered light intensity distribution data on the surface of the ore particles; The surface scattering characteristics of the ore particles are determined based on the scattered light intensity distribution data.

3. The method as described in claim 1, characterized in that, The trajectory offset coefficient of the ore particles generated based on the spatial contour features and the surface scattering features specifically includes: The equivalent volume and centroid coordinates of the ore particles are calculated based on the aforementioned spatial contour features. The compositional properties of ore particles are classified based on the surface scattering characteristics. The equivalent volume, the centroid coordinates, and the composition attributes are input into a preset trajectory offset model, and the initial trajectory offset coefficient of the ore particles is output. The initial trajectory offset coefficient is dynamically corrected based on the initial velocity of the ore particles to obtain the trajectory offset coefficient of the ore particles.

4. The method as described in claim 1, characterized in that, The prediction of ore particle trajectory partitioning using the trajectory offset coefficient and the local flow field parameters of the original ore flow specifically includes: Local flow field parameters of the raw ore flow within the sorting channel are collected, including airflow velocity vector and pressure distribution. The trajectory offset coefficient and the local flow field parameters are input into the particle motion dynamics model to calculate the motion acceleration of the ore particles under the action of airflow. Based on the aforementioned motion acceleration and the initial motion state of the ore particles, the complete motion trajectory of the ore particles is simulated and generated. The trajectory of the ore particles is divided into landing point zones based on the endpoint coordinates of the complete motion trajectory.

5. The method as described in claim 1, characterized in that, Mapping the landing point trajectory partitions to ore sorting windows specifically includes: Establish a spatial mapping relationship between the landing point trajectory partition and the physical coordinates of the sorting channel; Based on the spatial mapping relationship, determine the intervention plane area corresponding to the landing trajectory partition within the sorting channel; Based on the geometric boundary parameters of the intervention plane region, an ore sorting window is generated corresponding to the landing point trajectory partition.

6. The method as described in claim 1, characterized in that, Within the ore sorting window, generating airflow intervention commands based on the decision-making influence relationship of the trajectory offset coefficient on the landing point trajectory partition specifically includes: Establish the decision influence weight matrix of the trajectory offset coefficient and the landing point trajectory partition; The airflow intervention intensity level corresponding to the ore sorting trajectory offset coefficient is determined based on the decision influence weight matrix. The spatiotemporal parameter combination of the airflow pulse is generated based on the airflow intervention intensity level and the spatial position parameters of the ore sorting window; The spatiotemporal parameter combination is converted into a driving signal for a multi-level pulsed air curtain, forming an airflow intervention command.

7. The method as described in claim 1, characterized in that, In response to the airflow intervention command, determining the ore rejection airflow zone and the ore retention airflow zone in the ore particle sorting process specifically includes: Analyze the pressure gradient distribution in the airflow intervention command; The threshold boundary for the intensity of airflow disturbance within the sorting channel is defined based on the pressure gradient distribution. The region in the sorting channel where the pressure gradient value is greater than the threshold boundary is defined as the ore rejection airflow zone; The region in the sorting channel where the pressure gradient value is less than or equal to the threshold boundary is defined as the ore retention gas flow zone.

8. A smart ore pre-selection system based on airflow separation, used to execute the smart ore pre-selection method based on airflow separation as described in any one of claims 1 to 7, characterized in that, The system includes: The feeding module is used to uniformly feed the raw ore stream into the sorting channel through a vibrating feeder, so that the raw ore stream forms a controllable sparse distribution in the free fall area. The processing module is used to identify the spatial contour features and surface scattering features of ore particles upstream of the free fall zone, and generate the trajectory offset coefficient of the ore particles based on the spatial contour features and the surface scattering features. The processing module is also used to predict the landing trajectory partitions of ore particles by using the trajectory offset coefficient and the local flow field parameters of the original ore flow, and to map the landing trajectory partitions as ore sorting windows. Within the ore sorting window, airflow intervention commands are generated based on the decision influence relationship of the trajectory offset coefficient on the landing trajectory partitions. The processing module is also used to respond to the airflow intervention command, determine the ore rejection airflow zone and the ore retention airflow zone in the ore particle sorting process, and use a multi-stage pulse air curtain to send ore particles that meet the waste rock characteristics into the ore rejection airflow zone, while ore particles that do not meet the waste rock characteristics into the ore retention airflow zone. The execution module is used to acquire the landing point response signal of ore particles through the receiving end set in the ore rejection airflow zone and the ore retention airflow zone after the airflow intervention, and update the airflow intervention command for subsequent ore particle sorting with the landing point response signal.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the intelligent pre-selection method for ore based on airflow separation as described in any one of claims 1 to 7.

Images