Crusher abnormal state monitoring method and system based on image features

Abstract

The application belongs to the technical field of image processing, and particularly relates to a crusher abnormal state monitoring method and system based on image features, which comprises the following steps: acquiring a no-load image and a discharge image sequence of a discharge port, extracting foreground pixel points and active pixel points to obtain a material retention index; dividing the discharge port area into multiple sub-areas, and acquiring a discharge accumulation imbalance index according to local filling density and active density distribution; acquiring a real-time current load of a main shaft motor, combining the material retention index and the discharge accumulation imbalance index to obtain a power consumption form coordination factor; obtaining a material level accumulation trend parameter based on the change rate of the real-time material level height of a feeding port, combining the power consumption form coordination factor to obtain a crusher abnormal tendency degree, and determining whether the crusher enters an abnormal state. The application offsets normal high-yield overfeeding and environmental noise interference through cross-physical field data coordination, and improves the accuracy of equipment monitoring under complex working conditions.

Description

Technical Field

[0001] This invention relates to the field of image processing technology. More specifically, this invention relates to a method and system for monitoring abnormal conditions of a crusher based on image features. Background Technology

[0002] Crusher plays an important role in industrial automation processing fields such as mining and building materials. During the operation of crushers, abnormal problems such as material accumulation and blockage, equipment imbalance and mechanical overload often occur. Therefore, abnormal detection of crusher operation status is particularly important to ensure the physical operation safety of the equipment and the continuity of production.

[0003] In related technologies, for example, Chinese patent application document with publication number CN120410995A discloses a method, medium, and equipment for monitoring the status of a crusher based on changes in the morphology of particulate materials. The method includes: acquiring historical point cloud data of particulate materials in the crusher, extracting morphological parameters of the particulate materials, wherein the morphological parameters include the morphological parameters of each particle, the correlation parameters between particles, the morphological parameters of the material composed of particles, and the spatial parameters between the material and the crusher; constructing and training a monitoring model based on the crusher status marked by the historical point cloud data and the extracted parameters; and then obtaining the crusher status by extracting the current morphological parameters of the particulate materials and inputting them into the model.

[0004] However, in such harsh physical environments, the light or electromagnetic signals emitted by point cloud data acquisition equipment are easily subject to severe scattering and physical obstruction by dense dust, resulting in a large number of voids and noise in the acquired point cloud data. This related technology relies on the precise extraction of microscopic particle morphology and local spatial features, but under specific environmental interference from strong dust obstruction and severe physical vibration, it cannot stably acquire true three-dimensional particle structure data. Furthermore, this technology only analyzes the spatial morphology of the material, failing to coordinate with the equipment's own electrical load pressure and the overall physical throughput balance, leading to a high likelihood of false alarms under complex and harsh operating conditions. Summary of the Invention

[0005] To address the aforementioned technical problems of high misjudgment rate and inability to accurately identify true off-center load and blockage under strong dust interference and high-frequency dynamic disturbances, the present invention provides solutions in the following aspects.

[0006] In a first aspect, the present invention provides a method for monitoring abnormal states of a crusher based on image features, comprising: acquiring an unloaded image and a sequence of discharge images of the crusher discharge port; extracting the absolute value of the grayscale difference between corresponding positions of the images in the sequence and the unloaded image to determine foreground pixels; extracting the absolute value of the grayscale difference between corresponding positions of adjacent frames in the discharge image sequence to determine active pixels; calculating the ratio of the number of foreground pixels to the total number of pixels within the discharge port boundary in the unloaded image, and combining this with the ratio of the number of active pixels to the number of foreground pixels to obtain a material retention index; dividing the discharge port area into multiple sub-regions; determining the local filling density based on the number of foreground pixels in each sub-region; and... The active density distribution is determined based on the number of active pixels in each sub-region; the standard deviation of the local filling density of all sub-regions is calculated, and the discharge accumulation imbalance index is obtained by combining the average local filling density of all sub-regions and the sum of the active density distributions of all sub-regions; the instantaneous current value of the spindle motor is obtained, and the real-time current load is obtained based on the difference between the instantaneous current value and the reference current value under no-load conditions; the power consumption pattern coordination factor is obtained based on the material retention index, the discharge accumulation imbalance index, and the real-time current load; the crusher abnormality tendency is obtained based on the power consumption pattern coordination factor and the rate of change of the real-time material level height at the feed inlet; and the crusher abnormality tendency is used to determine whether the crusher has entered an abnormal state.

[0007] This invention extracts foreground and active pixels from the discharge port image, transforming the physical occlusion state and dynamic material flow rate of the crusher discharge port into a material retention index, thus eliminating the interference of normal high-output material flow on equipment blockage detection. Simultaneously, by dividing the discharge port into sub-regions and extracting local filling density and active density distributions, this invention transforms the eccentric feeding and local wear-induced off-center loading physical phenomena within the crushing chamber into spatially discrete characteristics of the discharge accumulation imbalance index, avoiding the risk of false detection caused by natural discharge bias. This invention also extracts the real-time current of the spindle motor. The load is combined with the stagnation and imbalance characteristics of the visual dimension to obtain the power consumption pattern coordination factor, which offsets the interference of visual data fluctuations caused by single ambient light or camera shake. The present invention combines the material level accumulation trend parameter of the feed inlet to transform the physical imbalance phenomenon of material intake and discharge into the adjustment weight of the crusher's abnormal tendency, so that the abnormal characteristics of the discharge end correspond synchronously with the congestion state of the feed end, thereby more accurately matching the actual working load of the crusher, which is conducive to improving the reliability of the crusher's abnormal state monitoring results and limiting the occurrence of engineering risks such as continuous cavity blockage and mechanical overload.

[0008] Preferably, the step of obtaining the crusher's abnormal tendency based on the power consumption pattern coordination factor and the rate of change of the real-time material level height at the feed inlet includes: obtaining the material level accumulation trend parameter based on the rate of change of the real-time material level height at the feed inlet, and obtaining the crusher's abnormal tendency based on the material level accumulation trend parameter and the power consumption pattern coordination factor.

[0009] Preferably, the abnormal tendency of the crusher satisfies the following relationship: In the formula, For the first The abnormal tendency of the crusher at each sampling time; For the first Power consumption pattern coordination factor corresponding to each sampling moment; For the first Material accumulation trend parameters corresponding to each sampling time; It is the hyperbolic tangent function; This is the trend adjustment coefficient.

[0010] This invention continuously acquires the real-time material level height at the feed inlet and calculates the average rising slope, transforming the physical accumulation of material at the feed end into a material level accumulation trend parameter. When the feed accumulates rapidly, it amplifies the impact of abnormal power consumption at the discharge end on the final tendency. When the feed is flat or decreasing, it softens the growth of abnormal data at the discharge end, coordinating the states at both ends to offset misjudgments caused by single feed fluctuations, which is beneficial for accurately identifying the real cavity blockage fault.

[0011] Preferably, the power consumption pattern coordination factor satisfies the following relationship: In the formula, For the first Power consumption pattern coordination factor corresponding to each sampling time. For the first Real-time current load corresponding to each sampling time. For the first The material retention index corresponding to the frame image. For the first The material stacking imbalance index corresponding to the frame image. It is an exponential function with the natural constant as its base. This is the nonlinear coupling gain coefficient.

[0012] This invention transforms the physical phenomena of material discharge stagnation and spatial load imbalance from the visual dimension into amplified weights for electrical load signals. When severe stagnation and imbalance occur at the discharge port, the impact of real-time current load on anomaly detection is amplified, transferring the visually adverse operating conditions to the increase in the power consumption coordination factor. When any indicator in the visual features indicates smooth material flow, the interference of occasional current data spikes on the overall result is reduced, decoupling false anomalies caused by single power grid fluctuations or single visual obstructions, which helps improve the reliability of heavy load condition detection.

[0013] Preferably, obtaining the real-time current load based on the difference between the instantaneous current value and the reference current value under no-load conditions includes: using the average current reading of the main shaft motor under no-load conditions as the reference current value; acquiring the real-time current sequence during the operation of the crusher, extracting the instantaneous current value of the main shaft motor at the sampling time; calculating the difference between the instantaneous current value and the reference current value, and calculating the ratio of the difference to the reference current value to obtain the real-time current load.

[0014] Preferably, the step of extracting the absolute value of the grayscale difference between the corresponding positions of the image in the sequence and the empty image to determine the foreground pixel includes: comparing the absolute value of the grayscale difference between the corresponding coordinate pixels of the extracted image in the sequence and the empty image with a preset background threshold pixel by pixel; and determining the pixel at the corresponding coordinate as the foreground pixel if the absolute value of the grayscale difference is greater than the background threshold.

[0015] Preferably, the material retention index satisfies the following relationship: In the formula, For the first The material retention index corresponding to the frame image. For the first The instantaneous material coverage pixel parameters corresponding to the frame image. Based on the pixel parameters of the discharge port, For the first Frame image and the first Inter-frame active pixel parameters corresponding to the frame image It is an exponential function with the natural constant as its base. This represents the retention sensitivity coefficient.

[0016] This invention uses the ratio of active pixels to foreground pixels as the basis for nonlinear adjustment, transforming the dynamic flow activity of material at the discharge port into a weighting factor for the total material coverage area. When the proportion of active pixels increases, the positive impact of excessive foreground pixels on blockage determination is weakened, offsetting the misleading effect of large-area physical occlusion caused by the equipment operating under high-volume and rapid discharge conditions. When the proportion of active pixels decreases, the weight of the foreground coverage area on retention assessment is amplified, enabling the final material retention index to reflect the risk of material stagnation and clogging, which is beneficial to improving the sensitivity of detecting the initial stage of minor blockages.

[0017] Preferably, determining the local fill density based on the number of foreground pixels in each sub-region includes: counting the number of foreground pixels falling into each sub-region to obtain the local fill pixel value corresponding to each sub-region; calculating the ratio of the local fill pixel value to the total number of foreground pixels, and using this ratio as the local fill density of the sub-region.

[0018] Preferably, the discharge stacking imbalance index satisfies the following relationship: In the formula, For the first The material stacking imbalance index corresponding to the frame image. For the first The first frame of the image Local fill density of each sub-region The standard deviation of the local fill density for all sub-regions. This represents the average local fill density across all sub-regions. For the first The sum of the activity density distributions of each sub-region in the frame image.

[0019] This invention transforms the uneven spatial distribution of crushed materials into basic off-center load discrete data by calculating the ratio of the standard deviation to the average value of the local filling density in a sub-region. Simultaneously, by introducing the sum of active density distributions, the overall material physical flow velocity at the discharge port is converted into a control weight for the degree of spatial dispersion. When the overall active density increases, it indicates a faster overall material flow velocity, thus weakening the impact of static uneven distribution on imbalance assessment. Conversely, when the overall active density decreases, it amplifies the impact of distribution dispersion on off-center load risk. This allows the discharge accumulation imbalance index to approximate the true off-center load state, improving the accuracy of local blockage detection.

[0020] Secondly, the present invention provides a crusher abnormal state monitoring system based on image features, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned crusher abnormal state monitoring method based on image features is implemented.

[0021] By adopting the above technical solution, the above-mentioned image feature-based crusher abnormal state monitoring method is generated into a computer program and stored in a memory for loading and execution by a processor. This allows for the creation of a terminal device based on the memory and processor, making it convenient to use.

[0022] The beneficial effects of this invention are as follows: This invention synchronously transforms the physical obstruction area of ​​the material at the discharge port and the dynamic displacement velocity into control weights for the material retention state, thus eliminating the interference of continuous physical obstruction caused by high-yield and rapid material flow; This invention transforms the distribution dispersion on the discharge cross-section into the characteristics of material accumulation imbalance, avoiding visual data misleading caused by local dead zones and limiting the risk of equipment wear caused by long-term local off-center loading; This invention transforms the visually captured material stagnation and spatial distribution off-center loading phenomenon into an amplified control factor for the real-time current load of the main shaft motor, offsetting the single data glitch caused by dust obstruction or power grid fluctuations, making the fused collaborative characteristics truly close to the physical extrusion pressure currently borne by the crusher main shaft; This invention extracts the continuous upward slope of the material level height at the feed port, transforming the physical imbalance phenomenon of upstream feeding far exceeding downstream discharge into a surge driving force for abnormal tendency. This operation synchronizes the physical throughput status at both ends, which not only softens false alarms caused by brief feeding fluctuations, but also quickly transfers the warning signal to the main control PLC for power-off and emergency stop actions when a substantial obstruction occurs, ensuring the physical operation safety of the equipment. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating the image feature-based crusher abnormal state monitoring method of the present invention; Figure 2 This is a schematic diagram illustrating the abnormal state monitoring of the crusher in this invention. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0026] This invention discloses a method for monitoring abnormal states of a crusher based on image features, referring to... Figure 1 This includes steps S1-S4: S1. Obtain the grayscale image sequence of the material discharge by using an industrial camera located at the discharge port, and combine it with the maximum load-bearing pixel parameter calibrated during the system's no-load self-test phase to obtain the material retention index.

[0027] It should be noted that the material throughput at the crusher discharge port typically fluctuates dynamically, and blockage is usually determined by the area of ​​material obstruction in the image. Since the discharged material is in a continuous accumulation state, a large obstruction area does not necessarily indicate a blockage; it may simply mean the equipment is in a normal high-output discharge state, leading to false alarms. Therefore, this invention uses an industrial camera located at the discharge port to acquire a sequence of grayscale images of the discharged material, and combines this with the maximum load-bearing pixel parameters calibrated during the system's no-load self-test phase to obtain the material retention index.

[0028] Specifically, under the condition of power-on initialization of the crusher and complete emptying of the discharge chute, a grayscale image of the discharge port area is acquired and marked as an empty image. The total number of pixels within the boundary of the discharge port in the empty image is used as the baseline pixel parameter of the discharge port. A sequence of discharge images during the crusher's operation is acquired. The first image in the sequence is extracted. The absolute value of the grayscale difference between corresponding pixels in the frame image and the idle image is used to compare the absolute value of the grayscale difference with a pre-defined background threshold pixel by pixel. If the absolute value of the grayscale difference is greater than the background threshold, the pixel corresponding to that coordinate is determined as a foreground pixel. The total number of pixels determined as foreground pixels is taken as the nth... The instantaneous material coverage pixel parameters corresponding to the frame image.

[0029] Furthermore, extract the first image from the discharge image sequence. Frame image and the first The absolute value of the grayscale difference between pixels at corresponding coordinate positions in a frame image, with the number of pixels whose absolute grayscale difference is greater than a preset fluctuation threshold as the number of pixels. Frame image and the first The inter-frame active pixel parameters corresponding to the frame image. Based on the baseline discharge port pixel parameters, instantaneous material coverage pixel parameters, and inter-frame active pixel parameters, obtain the first... The material retention index corresponding to the frame image.

[0030] For example, the background threshold is 15 and the preset fluctuation threshold is 35.

[0031] Specifically, the material retention index satisfies the following relationship: ; In the formula, For the first The material retention index corresponding to the frame image. For the first The instantaneous material coverage pixel parameters corresponding to the frame image. Based on the pixel parameters of the discharge port, For the first Frame image and the first Inter-frame active pixel parameters corresponding to the frame image It is an exponential function with the natural constant as its base. The retention sensitivity coefficient, The empirical range is In this embodiment for The implementers can determine the implementation based on the actual situation. When the crusher processes extremely fine manufactured sand with high moisture content leading to poor flowability, this parameter can be appropriately increased to improve sensitivity to minor stagnation phenomena. When processing large, coarsely crushed materials with violent jumping in the chute, this parameter can be appropriately decreased to prevent misjudgment caused by excessive local differences due to normal jumping.

[0032] in, This represents the instantaneous material coverage at the discharge port; The larger the value, the higher the proportion of the physical area of ​​the discharge port that is blocked by materials, the fuller the potential space for blockage, and the larger the base value of the material retention index. The smaller the value, the more open the discharge port is, resulting in a smaller base value for the material retention index; This represents the dynamic active percentage of the covered materials. The larger the value, the more drastic the material displacement or flow remains within the currently blocked area, and the less likely it is to stagnate or become blocked. This results in a smaller material retention index, which helps to eliminate interference from normal material flow during high production. The smaller the value, the more likely the material covering the discharge port is to be in a static state, increasing the likelihood of accumulation and blockage. The closer the index is to 1, the weaker its effect on reducing and inhibiting the material coverage area, thus resulting in a larger material retention index.

[0033] S2. Based on the instantaneous material coverage pixel parameters and inter-frame active pixel parameters, obtain the material discharge accumulation imbalance index through spatial distribution analysis.

[0034] It should be noted that the material distribution at the crusher discharge port directly reflects the load uniformity inside the crushing chamber. Typically, if localized wear or eccentric feeding occurs inside the crushing chamber, material will accumulate densely on one side of the discharge port while the other side remains empty, resulting in an unbalanced load. Therefore, this invention obtains the discharge accumulation imbalance index based on instantaneous material coverage pixel parameters and inter-frame active pixel parameters through spatial distribution variance analysis.

[0035] Specifically, the discharge port area is divided into Each sub-region is counted. The number of foreground pixels falling into each sub-region is counted to obtain the local padding pixel value corresponding to each sub-region. The local padding pixel value is then compared with the first sub-region. The ratio of the instantaneous material coverage pixel parameters corresponding to the frame image is the local fill density of the sub-region. Pixels within each sub-region that satisfy the condition that the absolute value of grayscale change is greater than a preset fluctuation threshold are extracted, their number is counted, and divided by the threshold. The active pixel parameters between frames corresponding to the frame image are used to obtain the active density distribution of each sub-region. The material discharge stacking imbalance index is obtained by combining the local fill density of all sub-regions with the sum of the active density distribution of each sub-region.

[0036] For example, the discharge port area is divided into Sub-regions.

[0037] Specifically, the material discharge and accumulation imbalance index satisfies the following relationship: ; In the formula, For the first The material stacking imbalance index corresponding to the frame image. For the first The first frame of the image Local fill density of each sub-region The standard deviation of the local fill density for all sub-regions. This represents the average local fill density across all sub-regions. For the first The sum of the activity density distributions of each sub-region in the frame image.

[0038] in, This reflects the unevenness of the discharge area. Typically, if the crushing chamber experiences eccentric feeding or localized wear, the crushed material will be concentrated on one side of the discharge port while remaining empty on the other. The larger the value, the higher the dispersion of the amount of crushed material between each sub-region, reflecting the greater possibility of the crusher operating under unbalanced load, resulting in a higher growth base for the discharge accumulation imbalance index, thus playing a basic role in characterizing anomalies.

[0039] pass right Corrections are made. A higher sum of active density distributions across all sub-regions indicates a faster overall material flow rate at the discharge port and a greater likelihood of no local dead zones, thus suppressing the discharge accumulation imbalance index and reflecting that while uneven distribution may exist, the material is still flowing rapidly. Conversely, a lower sum of active density distributions across all sub-regions indicates a greater likelihood of large-area static accumulation or localized blockages at the discharge port, leading to… The larger the value, the greater the impact of the distribution dispersion on the abnormal tendency of the crusher, reflecting the greater the possibility of severe off-center loading blockage.

[0040] S3. Based on the material retention index and the discharge accumulation imbalance index, combined with the real-time current load of the spindle motor, the power consumption mode coordination factor is obtained.

[0041] It should be noted that during normal operation of the crusher, the increase in output is accompanied by an increase in current. At this time, the material retention and discharge accumulation imbalance indices are within the normal fluctuation range, and this high power consumption and high flow is healthy. However, when mechanical blockage or structural damage occurs, high power consumption and low flow will appear. The conflict between these characteristics can be used to identify crusher malfunctions. Therefore, this invention derives a power consumption pattern coordination factor based on the material retention index and discharge accumulation imbalance index, combined with the real-time current load of the main shaft motor.

[0042] Specifically, using the average current reading of the main shaft motor under no-load conditions as the reference current value, the real-time current sequence during the operation of the crusher is obtained, and the first current is extracted. The sampling time (the first sampling time) The instantaneous current value at the time of frame image acquisition is recorded. The ratio of the difference between the instantaneous current value and the reference current value to the reference current value is calculated to obtain the real-time current load. The power consumption mode coordination factor is obtained based on the real-time current load, material retention index, and discharge stacking imbalance index.

[0043] Specifically, the power consumption pattern coordination factor satisfies the following relationship: ; In the formula, For the first Power consumption pattern coordination factor corresponding to each sampling time. For the first Real-time current load corresponding to each sampling time. For the first The material retention index corresponding to the frame image. For the first The material stacking imbalance index corresponding to the frame image. It is an exponential function with the natural constant as its base. For nonlinear coupling gain coefficients, The empirical range is In this embodiment for The implementers can determine the implementation based on the actual situation. When the material being processed by the crusher is extremely tough and prone to agglomeration, this parameter can be appropriately increased to amplify the effect of retention and imbalance on power consumption risk.

[0044] in, This represents the degree of co-deterioration of visual features; only when the first... When the frame image shows that the discharge port is both experiencing material stagnation and exhibiting severe distribution imbalance, Only then will significant values ​​be produced. This represents the crusher's performance in the [stage / phase]. The greater the probability that the discharge end at each sampling time will enter an extremely harsh working condition; The smaller the value, the better. The more likely the material is flowing rapidly at the discharge port corresponding to the frame image, or the more likely the material is distributed relatively evenly in each sub-region.

[0045] In particular, when the material retention index Or the imbalance index of material discharge accumulation When at least one of the indicators is at an extremely low level, even if the other indicator fluctuates slightly, It will remain at a very low value. This indicates that the crusher in the first... The higher the probability that the discharge end is in good operating condition at each sampling time, the more likely it is to lead to Its contribution to the power consumption pattern coordination factor is close to zero, thus suppressing power consumption fluctuations and effectively avoiding false alarms caused by single visual fluctuations.

[0046] Represents the first The basic power output performance of the crusher's main drive system at each sampling time. This item is calculated by... The difference between the instantaneous current value of the spindle motor and the reference current value at each sampling moment eliminates the inherent power consumption required by the motor to overcome mechanical friction and the inertia of the transmission system, thus reflecting the value of the first sampling moment. The load pressure generated in the crushing chamber due to the material crushing action at each sampling time. The larger the value, the more likely it is to be the first. The higher the intensity of the crushing operation being carried out by the crusher at a given sampling time, the greater the possibility that the mechanical structure is under excessive pressure. As the basis of power consumption pattern coordination factor, it provides a direct electrical indicator for assessing whether a device is in an overloaded or stall-prone state.

[0047] S4. Based on the power consumption mode coordination factor and combined with the real-time material level height, obtain the abnormal tendency of the crusher and make a status judgment.

[0048] It should be noted that the final health assessment of the crusher must take into account the feed-discharge balance. If abnormal power consumption occurs at the discharge end and the material level at the feed inlet shows a continuous stacking trend, it indicates that the crushing chamber has completely lost its processing capacity. Therefore, this invention uses a power consumption pattern coordination factor, combined with real-time material level height, to obtain the crusher's abnormal tendency and determine its status.

[0049] Specifically, the real-time material level height at the feed inlet is continuously acquired using radar sensors, and the average upward slope of the real-time material level height within a preset time window is calculated to obtain the material level accumulation trend parameter. The abnormal tendency of the crusher is then determined based on the power consumption pattern coordination factor and the material level accumulation trend parameter.

[0050] For example, the average rise slope of the real-time material level height over 30 seconds is calculated.

[0051] Specifically, the abnormal tendency of the crusher satisfies the following relationship: ; In the formula, For the first The abnormal tendency of the crusher at each sampling time; For the first Power consumption pattern coordination factor corresponding to each sampling moment; For the first Material accumulation trend parameters corresponding to each sampling time; It is the hyperbolic tangent function; This is the trend adjustment coefficient. The range of values ​​is In this embodiment, The implementers can adjust the system based on the inertial characteristics of the upstream feeding system. When the upstream feeding conveyor belt is long and there is a lot of residual material, the adjustment should be increased appropriately. This allows for a rapid increase in material prices should a slight upward trend emerge. This allows for earlier warning and shutdown to prevent cavity blockage; when the feeding system responds extremely quickly, the speed can be appropriately reduced. To improve production continuity.

[0052] in, It takes into account the material flow pattern at the discharge port, the dispersion of the accumulation distribution, and the real-time load pressure of the main motor. The larger the value, the more likely it is to be the first. The more obvious the abnormal characteristics of the crusher at the discharge end and power consumption end at each sampling time, the greater the possibility of substantial obstruction or overload of the equipment, thus making the final abnormality tendency greater. The smaller the value, the better. The frame image shows that the material discharge is smooth and evenly distributed, while the first... The fact that the spindle current at each sampling moment is within the range of no-load or light-load fluctuations reflects that the more stable the equipment's operating status, the lower the possibility of obstruction or failure, resulting in a lower tendency for the final crusher to malfunction, thus suppressing misjudgments caused by environmental interference or feed fluctuations.

[0053] This represents the trend of material accumulation at the feed end, reflecting the balance between output and input within the crushing chamber. When A positive value, and the larger the value, the greater the likelihood of material saturation and overflow inside the crushing chamber; when A negative value indicates that the crushing chamber is gradually emptying. When material accumulates rapidly, Significantly greater than 0, Rapidly surged, making Greater than 1, thus affecting the first Power consumption pattern co-factor corresponding to frame image An upward correction is made to indicate abnormalities at the discharge end and congestion at the feed end, thereby triggering a shutdown protection mechanism. When the material level remains constant or decreases, Less than or equal to 0, For negative values ​​or values ​​approaching 0, Less than or close to 1, thus affecting the power consumption pattern coordination factor. It plays a role in suppressing or maintaining the status quo, indicating that the current power consumption fluctuation has not yet caused an imbalance between the feed and discharge, and is within a controllable range, thereby ensuring the continuous operation capability of the equipment.

[0054] Furthermore, if the crusher's abnormal tendency exceeds a preset warning threshold, the crusher is determined to have entered an abnormal state. For example, the preset warning threshold is... .

[0055] In one embodiment, an emergency stop command is sent to the main control PLC of the crusher, and the upstream feeder is shut down simultaneously.

[0056] For example, Figure 2 This is a schematic diagram of the abnormal state monitoring of the crusher in this invention. As can be seen from the figure, as time goes by and the working conditions gradually deteriorate, the abnormal tendency of the crusher shows a gradual upward trend, reflecting the intensification of the equipment's load imbalance and power consumption abnormality. When the tendency exceeds the preset warning threshold for three consecutive sampling times, the system determines that the equipment is in an extremely high-risk state and immediately issues an emergency stop command. Subsequently, the equipment physically cuts off the power and enters the power outage protection period, thus realizing the fault detection and control of the crusher.

[0057] This invention also discloses an image feature-based crusher abnormal state monitoring system, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement the image feature-based crusher abnormal state monitoring method according to this invention.

[0058] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

Claims

1. A method for monitoring abnormal states of a crusher based on image features, characterized in that, include: Obtain the sequence of unloaded and discharged images from the crusher's discharge port; Extract the absolute value of the grayscale difference between corresponding positions in the image and the empty image in the sequence to determine the foreground pixel; Extract the absolute value of the grayscale difference at corresponding positions in adjacent frames of the material feeding image sequence to determine active pixels; The material retention index is obtained by calculating the ratio of the number of foreground pixels to the total number of pixels within the discharge port boundary in the unloaded image, and combining this with the ratio of the number of active pixels to the number of foreground pixels. The discharge port area is divided into multiple sub-regions. The local fill density is determined based on the number of foreground pixels in each sub-region. The active density distribution is determined based on the number of active pixels in each sub-region. The standard deviation of the local fill density of all sub-regions is calculated, and the discharge accumulation imbalance index is obtained by combining the average local fill density of all sub-regions and the sum of the active density distributions of all sub-regions. The instantaneous current value of the spindle motor is obtained, and the real-time current load is obtained based on the difference between the instantaneous current value and the reference current value under no-load conditions. The power consumption pattern coordination factor is obtained based on the material retention index, the discharge accumulation imbalance index, and the real-time current load. The abnormal tendency of the crusher is obtained based on the power consumption mode synergy factor and the rate of change of the real-time material level height at the feed inlet. Determine whether the crusher has entered an abnormal state based on the degree of abnormality in the crusher.

2. The method for monitoring abnormal states of a crusher based on image features according to claim 1, characterized in that, The method of obtaining the crusher's abnormal tendency based on the power consumption pattern coordination factor and the rate of change of the real-time material level height at the feed inlet includes: obtaining the material level accumulation trend parameter based on the rate of change of the real-time material level height at the feed inlet, and obtaining the crusher's abnormal tendency based on the material level accumulation trend parameter and the power consumption pattern coordination factor.

3. The method for monitoring abnormal states of a crusher based on image features according to claim 2, characterized in that, The abnormal tendency of the crusher satisfies the following relationship: ； In the formula, For the first The abnormal tendency of the crusher at each sampling time; For the first Power consumption pattern coordination factor corresponding to each sampling moment; For the first Material accumulation trend parameters corresponding to each sampling time; It is the hyperbolic tangent function; This is the trend adjustment coefficient.

4. The method for monitoring abnormal states of a crusher based on image features according to claim 1, characterized in that, The power consumption mode coordination factor satisfies the following relationship: ； In the formula, For the first Power consumption pattern coordination factor corresponding to each sampling time. For the first Real-time current load corresponding to each sampling time. For the first The material retention index corresponding to the frame image. For the first The material stacking imbalance index corresponding to the frame image. It is an exponential function with the natural constant as its base. This is the nonlinear coupling gain coefficient.

5. The method for monitoring abnormal states of a crusher based on image features according to claim 1, characterized in that, The step of obtaining the real-time current load based on the difference between the instantaneous current value and the reference current value under no-load conditions includes: using the average current reading of the main shaft motor under no-load conditions as the reference current value; acquiring the real-time current sequence during the operation of the crusher, extracting the instantaneous current value of the main shaft motor at the sampling time; calculating the difference between the instantaneous current value and the reference current value, and calculating the ratio of the difference to the reference current value to obtain the real-time current load.

6. The method for monitoring abnormal states of a crusher based on image features according to claim 1, characterized in that, The step of extracting the absolute value of the grayscale difference between the corresponding positions of the image in the sequence and the empty image to determine the foreground pixel includes: comparing the absolute value of the grayscale difference between the corresponding coordinate pixels of the extracted image in the sequence and the empty image with a preset background threshold pixel by pixel; and determining the pixel at the corresponding coordinate as the foreground pixel if the absolute value of the grayscale difference is greater than the background threshold.

7. The method for monitoring abnormal states of a crusher based on image features according to claim 1, characterized in that, The material retention index satisfies the following relationship: ； In the formula, For the first The material retention index corresponding to the frame image. For the first The instantaneous material coverage pixel parameters corresponding to the frame image. Based on the pixel parameters of the discharge port, For the first Frame image and the first Inter-frame active pixel parameters corresponding to the frame image It is an exponential function with the natural constant as its base. This represents the retention sensitivity coefficient.

8. The method for monitoring abnormal states of a crusher based on image features according to claim 1, characterized in that, The step of determining the local fill density based on the number of foreground pixels in each sub-region includes: counting the number of foreground pixels falling into each sub-region to obtain the local fill pixel value corresponding to each sub-region; calculating the ratio of the local fill pixel value to the total number of foreground pixels, and using this ratio as the local fill density of the sub-region.

9. The method for monitoring abnormal states of a crusher based on image features according to claim 1, characterized in that, The discharge accumulation imbalance index satisfies the following relationship: ； In the formula, For the first The material stacking imbalance index corresponding to the frame image. For the first The first frame of the image Local fill density of each sub-region The standard deviation of the local fill density for all sub-regions. This represents the average local fill density across all sub-regions. For the first The sum of the activity density distributions of each sub-region in the frame image.

10. A crusher abnormal state monitoring system based on image features, characterized in that, include: A processor and a memory, wherein the memory stores computer program instructions that, when executed by the processor, implement the image feature-based crusher abnormal state monitoring method according to any one of claims 1-9.

Images