Machining oil mist separation real-time monitoring system

Abstract

The invention relates to the technical field of oil mist separation, in particular to a machining oil mist separation real-time monitoring system which comprises a particle trapping monitoring module, an airflow adjustment control module, a particle path judgment module, a boundary recognition analysis module and a separation efficiency feedback module. According to the method, through combined monitoring of the oil mist particle concentration and the airflow speed, dynamic analysis of the particle trapping efficiency and deviation direction identification are achieved, the diffusion trend identification precision is improved, the guide angle is introduced to be compared with the wind speed parameter, the insufficient response area is defined, the adjustment precision and real-time performance are enhanced, and the particle sedimentation direction is classified and identified; the method comprises the steps of constructing a structured motion path, locking space abnormal behaviors, fusing visible light and near-infrared refraction angle changes, identifying particle boundary features, improving the abnormal distribution analysis capability, carrying out superposition judgment on boundary abnormality and an adjustment response region, evaluating the separation stability, and feeding back separation capability distribution in a monitoring region. And the trend tracking and separation efficiency analysis capability is enhanced.

Description

Technical Field

[0001] The invention relates to the technical field of oil mist separation, and in particular to a real-time monitoring system for machining oil mist separation. Background Art

[0002] Oil mist separation technology is a key environmental control technology in the field of industrial manufacturing, and is mainly used in industrial scenarios such as machining and CNC machining. This technology targets the oil mist pollution problem generated by coolants or lubricants during high-speed cutting and grinding. Through physical means such as mechanical separation, centrifugal separation, electrical separation and filter materials, it realizes the whole process from oil mist collection, flow guidance, particle separation to purification and emission, effectively reducing the impact of oil mist on the production environment and the health of operators. It is an important part of the modern industrial clean production management system.

[0003] Among them, the machining oil mist separation real-time monitoring system refers to a system used to monitor and manage the operating status of the oil mist separation equipment during CNC machining, and is specifically used to collect and analyze key parameters such as the operating status and separation efficiency of the oil mist collection device. The system collects data such as changes in oil mist concentration, air flow velocity inside the equipment, and particle aggregation in the separation chamber through sensor devices, and judges the working status of the equipment based on the collected results. The completion method includes the deployment of specific collection structures such as optical detection components, gas flow rate detection units, and particle recognition devices. By collecting and analyzing various types of monitoring data, real-time grasp of the working status of the oil mist separation equipment and early warning of operating trends can be achieved.

[0004] However, the existing oil mist separation monitoring technology still has the following technical defects or deficiencies. In terms of data acquisition, the existing technology in oil mist separation monitoring mostly relies on single-point parameter acquisition methods, lacks continuous tracking of the dynamic path of oil mist diffusion and concentration trend changes, can only provide static data of the local state of the separation device, and is difficult to support the real-time feedback mechanism required for multi-point linkage control. In the airflow adjustment link, the existing technology mostly adopts a preset response strategy, ignoring the dynamic impact of oil mist concentration fluctuations on fan guide adjustment, resulting in a decrease in control efficiency and adjustment lag. In terms of particle identification, the existing technology is limited to quantitative statistics of particle concentration, lacks the ability to multi-dimensionally identify the direction of particle movement, sedimentation path and offset trend, and thus affects the overall judgment of particle diffusion behavior. In addition, the existing technology relies on simple threshold judgment in boundary anomaly identification, which cannot accurately reflect the subtle changes in particle optical characteristics and limits the ability to identify the spatial distribution of boundary anomalies. For example, in the multi-particle aggregation area, there are omissions or misjudgments in identification, which affects the system's accurate feedback on the change in separation efficiency, resulting in the inability to achieve dynamic tuning of the oil mist treatment effect under complex working conditions. Summary of the invention

[0005] The purpose of the present invention is to solve the defects or shortcomings in the prior art and to propose a real-time monitoring system for machining oil mist separation.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] The machining oil mist separation real-time monitoring system includes:

[0008] The particle capture monitoring module collects the oil mist particle concentration and air flow velocity of the CNC spindle exhaust channel and tool change position based on the machining oil mist separation monitoring data, analyzes the change of particle capture efficiency at the collection point, determines the offset from the concentration reference value, identifies the offset direction, and generates a particle concentration trend label at the monitoring position;

[0009] an airflow adjustment control module, based on the particle concentration trend tag of the monitoring position, identifying the fan orientation of the concentration rising area, determining whether the airflow adjustment response is lower than the fan linkage adjustment reference value, marking the position that does not meet the response standard, and outputting a fan response status list;

[0010] The particle path judgment module calls the fan response state list, identifies the particle settling direction trajectory in the airflow path, performs direction classification on the particle offset direction, records the identification information of the differentiated directional offset point, and obtains the particle path offset identification;

[0011] The boundary recognition and analysis module calls the sampling points located in the particle path offset mark, collects the visible light and near-infrared refraction angles under the incident path of the light beam, analyzes the changes in the incident and refraction angles when the light beam penetrates the particles, classifies and stores the boundary feature anomaly points, and obtains the boundary recognition anomaly distribution data.

[0012] As a further solution of the present invention, the particle concentration trend label of the monitoring position includes the concentration change amplitude, change trend type, and abnormal fluctuation frequency; the fan response status list includes response delay parameters, guide angle offset, and wind speed adjustment amplitude; the particle path offset identifier includes particle sedimentation direction, offset path characteristics, and directional classification number; the boundary identification abnormal distribution data includes optical refraction difference, abnormal boundary density, and refraction anomaly distribution range.

[0013] As a further solution of the present invention, the particle capture monitoring module includes:

[0014] The particle concentration monitoring submodule extracts the air flow velocity, particle concentration, channel cross-sectional area and sampling time of the CNC spindle exhaust channel and tool change position based on the machining oil mist separation monitoring data, identifies the particle throughput per unit time, analyzes the particle distribution differences within the same time period, and obtains the particle transfer volume change range;

[0015] The offset identification submodule compares the difference between the particle transfer amount change interval and the particle concentration change reference value, and combines the particle concentration change speed of the monitoring point with the offset rate mark of the transfer interval, using the formula: Calculate the particle concentration offset intensity value and obtain the concentration offset characteristic value;

[0016] Where ΔC represents the particle concentration offset intensity value, v i represents the airflow velocity at the i-th monitoring point, q i represents the particle concentration at the i-th monitoring point, B i Represents the baseline value of particle concentration change at the i-th monitoring point, A i represents the channel cross-sectional area of ​​the i-th monitoring point, t i represents the sampling time series index of the i-th monitoring point, and n represents the total number of monitoring points;

[0017] The trend label identification submodule calls the concentration offset feature, determines the offset progressive direction and accumulation, matches the trend extreme point coordinates, and generates a particle concentration trend label at the monitoring location.

[0018] As a further solution of the present invention, the airflow adjustment control module includes:

[0019] The wind direction labeling submodule identifies the area with increasing concentration based on the particle concentration trend label of the monitoring position, selects the fan numbers with large steering angles and low wind speeds, and generates a fan steering label set;

[0020] The response judgment submodule calls the number in the fan guide tag set to identify the corresponding real-time guide angle and wind speed value, using the formula: Calculate the airflow response offset value, filter the numbers whose offset exceeds the limit, and obtain the number set of non-compliant responses;

[0021] Wherein, R represents the airflow response offset value, α represents the real-time steering angle, D represents the set steering angle reference value, w represents the real-time wind speed value, L represents the distance between the steering point and the fan outlet, τ represents the response monitoring delay sequence number, and θ represents the fan adjustment reference angle difference;

[0022] The status output submodule combines the fan numbers and corresponding response statuses according to the response non-standard number set, sets the response status labels, enters all the corresponding relationships between numbers and statuses into the fan status array, and outputs the fan response status list.

[0023] As a further solution of the present invention, the particle path judgment module includes:

[0024] The path trajectory extraction submodule calls the guide angle and the spatial position coordinates of the monitoring point in the fan response state list, selects the sectors whose projection difference exceeds the set threshold, analyzes the degree of coincidence between the particle concentration distribution and the wind direction in the path section, and obtains the sedimentation direction trajectory sequence;

[0025] The directional classification submodule extracts the offset angle, monitoring point coordinate set and trajectory boundary parameters based on the settlement direction trajectory sequence, and classifies the trajectory according to the wind direction reference angle and spatial distribution, using the formula: Calculate the particle direction offset value, make direction attribution judgment, and generate the particle offset direction classification result;

[0026] Among them, Ψ represents the particle direction offset value, represents the trajectory deviation angle, E represents the dominant wind direction angle of the area, S1 represents the horizontal projection distance from the starting point of the trajectory to the monitoring point, S2 represents the vertical projection distance from the end point of the trajectory to the monitoring point, μ represents the trajectory number sequence value, ε represents the projection layer offset, η represents the number of sectors crossed by the trajectory, and δ represents the average trajectory deviation angle;

[0027] The offset identification recording submodule calls the particle offset direction classification result, filters the spatial identification points with inconsistent directional labels, matches the corresponding path trajectory number and the monitoring time period, marks the spatial coordinates and sequence number of the point, and obtains the particle path offset identification.

[0028] As a further solution of the present invention, the boundary recognition and analysis module includes:

[0029] The sampling optical monitoring submodule collects incident path data of the light beam in the visible light and near infrared bands based on the sampling points located in the particle path deviation mark, monitors the initial values ​​and time series changes of the incident angle and refraction angle at the sampling points, and obtains a group of incident and refraction angle changes;

[0030] The refraction change calculation submodule extracts the deflection trajectory and particle size width change value in the path of the light beam penetrating the particle according to the incident refraction angle change group, and performs cross-time series weighted accumulation using the formula: Calculate the boundary bending strength value to obtain the boundary recognition interference index;

[0031] Among them, Δθ represents the boundary bending strength value, Q a represents the change in the incident angle of visible light at the ath sampling position, β a represents the change in near-infrared refraction angle at the ath sampling position, F a represents the deflection length of the penetration trajectory at the ath sampling position, γ a represents the particle size path offset width at the ath sampling position, λ arepresents the beam propagation timing number of the ath sampling position, and z is the total number of sampling positions;

[0032] The abnormal boundary aggregation submodule calls the boundary identification interference index, extracts the spatial distribution of nodes whose values ​​are higher than the interference threshold, aggregates the label boundary data according to the value and position, and obtains the boundary identification abnormal distribution data.

[0033] As a further solution of the present invention, the system also includes a separation efficiency feedback module:

[0034] A separation efficiency feedback module determines whether there is an overlapping relationship between the distribution density and the fan response state based on the coverage number of the distribution points in the boundary recognition abnormal distribution data within the fan adjustment area, marks the number of the overlapping area, and outputs the separation stability distribution of the monitoring area;

[0035] The monitoring area separation stability distribution includes overlapping area identification, stability level classification, and density coverage ratio.

[0036] As a further solution of the present invention, the separation efficiency feedback module includes:

[0037] The boundary point analysis submodule identifies abnormal distribution data based on the boundary, identifies distribution points that meet the conditions, and determines whether the set boundary conditions are met according to the spatial position of each distribution point, calibrates the distribution points that meet the conditions, and generates distribution point boundary information;

[0038] A fan adjustment state identification submodule analyzes the coverage quantity of the distribution points in the fan adjustment area based on the distribution point boundary information, and compares it with the fan response state to obtain fan response adjustment state data;

[0039] The regional stability calibration submodule analyzes the overlapping relationship between the fan adjustment area and the distribution point density according to the fan response adjustment state data, calibrates the overlapping area number, and outputs the monitoring area separation stability distribution.

[0040] Compared with the prior art, the advantages and positive effects of the present invention are:

[0041] In the present invention, by jointly monitoring the oil mist particle concentration and air flow velocity during machining, dynamic analysis of particle capture efficiency can be achieved, so as to accurately identify the offset direction between the concentration change and the reference value, and significantly improve the identification accuracy of particle diffusion trend. In the concentration rising area, the fan guide angle and wind speed adjustment parameters are introduced for comparison and identification, and the air flow control area that fails to respond effectively is clearly identified, which significantly enhances the pertinence and dynamic response capability of the fan adjustment behavior. The sedimentation direction of the particles in the air flow path is classified and identified, forming a structured expression of the particle movement path, so that the abnormal behavior of the particles in the spatial position can be effectively locked. Through the analysis of the change in the refraction angle of the light beam, the visible light and near-infrared information are compositely processed to effectively realize the optical recognition of the particle boundary characteristics, and improve the spatial resolution capability of the abnormal distribution of the boundary, and further make a coincidence judgment between the abnormal distribution data of the boundary and the fan adjustment response area, accurately evaluate the stability performance of the particle separation behavior under dynamic airflow conditions, and finally form a spatial distribution feedback of the stable separation capability of the oil mist particles in the monitoring area. Starting from multi-dimensional perception parameters, a data-driven high-precision trend tracking and separation efficiency analysis chain is built to comprehensively improve the accuracy of monitoring results, the sophistication of control strategies and the stability assessment capabilities of separation systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 is a system flow chart of the present invention;

[0043] Figure 2 This is a flow chart of the particle capture monitoring module in the present invention;

[0044] Figure 3 This is a flow chart of the airflow adjustment control module in the present invention;

[0045] Figure 4 This is a flow chart of the particle path judgment module in the present invention;

[0046] Figure 5 This is a flow chart of the boundary identification and analysis module in the present invention;

[0047] Figure 6 This is a flow chart of the separation efficiency feedback module in the present invention. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0049] See also Figure 1 , the machining oil mist separation real-time monitoring system includes:

[0050] The particle capture monitoring module is based on the machining oil mist separation monitoring data, including the oil mist particle concentration and air flow velocity of the CNC spindle exhaust channel and the tool change position. It analyzes the changes in the oil mist particle capture efficiency at the collection point, determines the offset between the particle concentration change reference value, and identifies the offset direction, generating a particle concentration trend label at the monitoring position;

[0051] The airflow adjustment control module identifies the fan guide in the area of ​​rising concentration based on the particle concentration trend label at the monitoring location, calls the guide angle and wind speed parameters, determines whether the airflow adjustment response is lower than the fan linkage adjustment reference value, marks the location that does not meet the response standard, and outputs a fan response status list;

[0052] The particle path judgment module calls the fan response state list, identifies the particle settling direction trajectory in the airflow path, and directional classifies the particle offset direction, records the identification information of the differentiated directional offset point, and obtains the particle path offset identification;

[0053] The boundary recognition and analysis module calls the sampling points located in the particle path offset mark, collects the visible light and near-infrared refraction angles under the incident path of the light beam, analyzes the changes in the incident and refraction angles of the corresponding light beam when penetrating the particle, classifies and stores the boundary feature anomaly points, and obtains the boundary recognition anomaly distribution data;

[0054] The separation efficiency feedback module determines whether there is an overlapping relationship between the distribution density and the fan response state based on the coverage number of distribution points in the abnormal distribution data in the boundary recognition within the fan adjustment area, marks the number of the overlapping area, and outputs the separation stability distribution of the monitoring area.

[0055] The particle concentration trend label at the monitoring location includes the concentration change amplitude, change trend type, and abnormal fluctuation frequency. The fan response status list includes the response delay parameter, guide angle offset, and wind speed adjustment amplitude. The particle path offset identification includes the particle sedimentation direction, offset path characteristics, and directional classification number. The boundary identification anomaly distribution data includes the optical refraction difference, abnormal boundary density, and refraction anomaly distribution range. The monitoring area separation stability distribution includes the overlapping area identification, stability level division, and density coverage ratio.

[0056] See also Figure 2 , the particle capture monitoring module includes:

[0057] The particle concentration monitoring submodule, based on the machining oil mist separation monitoring data, extracts the air flow velocity, particle concentration, channel cross-sectional area and sampling time of the CNC spindle exhaust channel and the tool change position, identifies the particle throughput per unit time, analyzes the particle distribution differences within the same time period, and obtains the particle transfer amount change range.

[0058] The concentration and airflow velocity of the oil mist particles generated by the CNC machine tool during the machining process are continuously monitored by sensors, so as to calculate the total amount of particles passing through the CNC spindle exhaust channel and the tool change position under different working conditions, and the data is transmitted to the processing center in real time through data acquisition. By analyzing the data of different time periods, the time series changes of particle concentration at each monitoring point are compared. The working state of the machine tool is taken into account during the implementation process. For example, the change of particle concentration will be significantly different when cutting at high speed and cutting at low speed. This is clearly visible through real-time data monitoring. For example, when cutting at high speed, the concentration will increase sharply in a short period of time. Through detailed data analysis, the change range of particle transfer amount can be accurately obtained, and then the efficiency of machine tool exhaust and its performance under different working conditions can be judged.

[0059] The offset identification submodule compares the difference between the particle transfer amount change interval and the particle concentration change reference value, combines the particle concentration change speed of the monitoring point with the offset rate mark of the transfer interval, and uses the formula: Calculate the particle concentration offset intensity value and obtain the concentration offset feature value.

[0060] Where ΔC represents the particle concentration offset intensity value, v i represents the airflow velocity at the i-th monitoring point, q i represents the particle concentration at the i-th monitoring point, B i Represents the baseline value of particle concentration change at the i-th monitoring point, A i represents the channel cross-sectional area of ​​the i-th monitoring point, t i Represents the sampling time series index of the i-th monitoring point, and n represents the total number of monitoring points.

[0061] The monitoring data is compared with the particle concentration change reference value to further obtain the deviation direction and deviation amplitude of the particle concentration. i The setting of is based on the 5-minute average value collected by the equipment under stable operation, and the average value is set as the critical reference value of concentration deviation. Taking the tool change position as an example, the particle concentration fluctuation range under normal cutting state is 18-22, and its concentration change reference value B is set. i is 20, and the particle concentration q at each monitoring point i The data is obtained by a laser scattering particle counter with a sampling frequency of once every 10 seconds. The data of 10 consecutive sampling points are taken during the analysis period, corresponding to the air flow velocity v i The actual value measured by the hot film anemometer shall prevail. The sensors are placed in the spindle exhaust duct and tool change position, and the typical air flow velocity measured is 2.5, and the channel cross-sectional area A i Calculated based on the exhaust pipe diameter, the diameter is 80mm, then A i =π·(0.04) 2=0.0050, sampling time index t i represents the sampling sequence number. If the current round of calculation is the third sampling point, then t i =3;

[0062] Taking the number of monitoring points n=3 as an example, the data of the three monitoring points are as follows:

[0063] Monitoring point 1: q1=25, v1=2.3, B1=20, A1=0.0050, t1=3;

[0064] Monitoring point 2: q2=18, v2=2.6, B2=20, A2=0.0050, t2=3;

[0065] Monitoring point 3: q3=21, v3=2.5, B3=20, A3=0.0050, t3=3;

[0066] Substituting into the formula:

[0067] Calculate each term:

[0068] Monitoring point 1:

[0069] Monitoring points 2 items:

[0070] 3 monitoring points:

[0071] Sum and take the absolute value: ΔC = |0.6628+0.4737+0.5745| = 1.711;

[0072] The result shows that the combined particle concentration offset intensity value of the three monitoring points in the current sampling period is 1.711. This value can be compared with the empirically set offset intensity critical value of 2.0 to determine whether it is within the allowable variation range.

[0073] By introducing the joint calculation of multi-point concentration, velocity, and channel area, and normalizing them with the numbers of different sampling periods, the results can comprehensively reflect the overall trend and deviation degree of particle offset intensity at different monitoring points, thereby improving the sensitivity and accuracy of anomaly identification.

[0074] The trend label identification submodule calls the concentration offset feature, determines the offset progressive direction and accumulation, matches the trend extreme point coordinates, and generates the particle concentration trend label of the monitoring location.

[0075] Through in-depth analysis of the offset feature quantity, combined with real-time monitoring data, the overall trend of particle concentration can be determined. For example, through trend analysis software, data points can be connected to form a trend line of the time series. Through the trend line, you can intuitively see the upward or downward trend of particle concentration, and by setting a specific threshold. When the trend reaches the threshold, an alarm is automatically generated and the extreme point of the trend is marked. Marking helps operators quickly identify potential problem areas so that necessary adjustments or maintenance can be made. In this way, the air quality in the factory can be monitored and adjusted in real time to ensure the safety and cleanliness of the working environment, and particle concentration trend labels for the monitoring location are generated. The labels provide a basis for subsequent maintenance and optimization.

[0076] See also Figure 3 , the air flow adjustment control module includes:

[0077] The wind direction labeling submodule identifies the area with increasing concentration based on the particle concentration trend label at the monitoring location, selects the fan numbers with large steering angles and low wind speeds, and generates a fan steering label set.

[0078] Through monitoring, we can identify which areas are experiencing rising particle concentrations, and the data is obtained by comparing with historical data. For example, if the particle concentration in a certain area continues to rise and exceeds the set threshold for five consecutive minutes, it is marked as an area of ​​rising concentration, and the fans in the area need to be adjusted. The fan's steering angle and wind speed are automatically calculated and adjusted based on the fan's position and working environment. The adjustment standard is that the fan's steering angle needs to be greater than the set angle standard, and the wind speed needs to be lower than the set speed threshold, to ensure that the adjustment can effectively target the area with rising particle concentration, thereby reducing the particle concentration in the area and generating a fan steering label set, which contains the numbers of all fans that need to be adjusted and their new steering angles and wind speeds.

[0079] The response judgment submodule calls the number in the fan guide tag set to identify the corresponding real-time guide angle and wind speed value using the formula: Calculate the airflow response offset value, filter the numbers whose offset exceeds the limit, and obtain the set of numbers whose responses do not meet the limit.

[0080] Wherein, R represents the airflow response offset value, α represents the real-time steering angle, D represents the set steering angle reference value, w represents the real-time wind speed value, L represents the distance between the steering point and the fan outlet, τ represents the response monitoring delay number, and θ represents the fan adjustment reference angle difference.

[0081] Fans with significant deviations in steering angle and wind speed are selected for analysis. The current steering angle α and wind speed w of each fan are extracted from real-time monitoring and compared with the preset reference steering angle D and wind speed threshold. The preset value is set according to the fan position and the expected wind direction effect. For example, the fan steering angle reference value D is set to 45 degrees and the wind speed threshold is set to 2.0m / s. If the real-time monitoring shows that the steering angle α of a fan is 50 degrees and the wind speed w is 1.8m / s, the deviation is calculated.

[0082] Wherein, the distance L is the straight-line distance from the fan to the nearest concentration monitoring point, which is assumed to be 10 meters. The response time serial number τ represents the time delay from the adjustment command to the actual measurement, which is assumed to be the response time serial number within 5 minutes, and the average value is 3. The preset adjustment reference angle difference θ is preset according to the fan efficiency and the required coverage area size, and is set to 5 degrees. After substituting the numerical value, the formula calculation process is as follows:

[0084] This value R is the response offset value of the fan, which indicates the degree of difference between the actual response of the fan and the expected response. This value is compared with the preset response offset limit (for example, 3.5). If the R value exceeds 3.5, the fan is marked as not responding well and needs maintenance or adjustment. In this example, the R value is 1.2025, which does not exceed the limit, so the fan status is normal.

[0085] This analysis process not only helps identify fans that need maintenance, but also ensures that the fan adjustment strategy is consistent with the needs in the actual factory environment, thereby improving the efficiency and responsiveness of the entire system. In this way, the operating status of the fans can be effectively managed and the air quality control of the factory can be optimized.

[0086] The status output submodule combines the fan number and the corresponding response status according to the response non-standard number set, sets the response status label, enters the corresponding relationship between all numbers and status into the fan status array, and outputs the fan response status list.

[0087] Mark the current status of each numbered fan as "abnormal" or "normal", and this marking process is automatic. Compare the real-time data of each fan with the set response threshold. If the data shows that the fan's response offset exceeds the threshold, it is marked as "abnormal", otherwise it is marked as "normal". In this way, operators can clearly know which fans need attention and which are working normally. This not only helps maintenance personnel quickly locate problem fans, but also helps to perform systematic maintenance and optimization. Integrate all fan numbers and their corresponding status into a list, and output the fan response status list. This list can be directly used for subsequent monitoring and maintenance work to ensure the overall operating efficiency of the system.

[0088] See also Figure 4 , the particle path judgment module includes:

[0089] The path trajectory extraction submodule calls the guidance angle and the spatial position coordinates of the monitoring point in the fan response status list, selects the sectors whose projection difference exceeds the set threshold, analyzes the degree of coincidence between the particle concentration distribution and the wind direction in the path section, and obtains the sedimentation direction trajectory sequence.

[0090] Determine whether the current status of each fan is normal, and perform preliminary screening based on this status to select fans with abnormal status. For example, if the response status of a fan deviates significantly from the predetermined normal operating parameters, such as the guide angle or wind speed exceeds the normal operating range, the fan will be marked as abnormal. Use spatial position coordinate data to analyze the specific location of the abnormal fan and the affected particle motion trajectory. Calculate the particle settling area by simulating the new path of the particles after the fan is adjusted. By comparing the simulation results with the actual monitoring data, confirm which areas are significantly affected by the particle concentration of the fan status. Consider the specific fault type of the fan, the distance between the fan and the monitoring point, and the physical characteristics of the particles such as size and weight, perform detailed calculations, and generate a settling direction trajectory sequence. This sequence provides a dynamic view of the potential settling path of particles corresponding to each fan.

[0091] The directional classification submodule extracts the offset angle, monitoring point coordinate set and trajectory boundary parameters based on the settlement direction trajectory sequence, and classifies the trajectory according to the wind direction reference angle and spatial distribution, using the formula:

[0092] Calculate the particle direction offset value, make a direction attribution judgment, and generate the particle offset direction classification result.

[0093] Among them, Ψ represents the particle direction offset value, represents the trajectory deviation angle, E represents the dominant wind direction angle of the area, S1 represents the horizontal projection distance from the starting point of the trajectory to the monitoring point, S2 represents the vertical projection distance from the end point of the trajectory to the monitoring point, μ represents the trajectory number sequence value, ε represents the projection layer offset, η represents the number of sectors crossed by the trajectory, and δ represents the average trajectory deviation angle;

[0094] Further determine the directional attribution of particles in space. First, extract the spatial offset angle of each trajectory This angle is calculated from the angle between the vector formed from the starting point to the end point of the particle path and the main wind direction vector. The main wind direction angle E is calculated by counting the direction pointed by the maximum number of guide angles in the fan response state list. This direction is regarded as the dominant airflow path direction in the wind field. To facilitate subsequent calculations, all angle parameters are uniformly expressed in degrees. Subsequently, the horizontal projection distance S1 and the vertical projection distance S2 between the two end points of the trajectory to the monitoring point are measured. Using the three-dimensional space coordinate projection method, S1 and S2 are calculated in meters, and are unfolded in the horizontal plane and the vertical plane respectively as Euclidean distance. The trajectory sequence value μ is obtained by numbering the trajectory in sequence in the time dimension, which is an integer serial number. ε is the hierarchical projection offset item corresponding to the number, which is used to correct the overlapping projections of different airflow levels. η represents the number of sectors crossed in the particle path, which is obtained by the angle segmentation statistics of the fan response section. δ is the average value of the direction angles of all nodes in the trajectory path, which is obtained by calculating the arithmetic mean of the direction angle set. It is a unified unit and is the same as E remains consistent.

[0095] The assignment is as follows: (path direction angle), E = 40 (main wind direction angle), S1 = 6.25 (meters), S2 = 2.25 (meters), μ = 4 (track number), ε = 1 (track layer number correction term), η = 3 (number of sectors crossed by the path), δ = 10 (path average deviation angle);

[0096] Substitute the formula into the following calculation:

[0097] The final calculated value of the particle direction offset Ψ is 10.493. This value is compared with the set classification boundary value. For example, if the maximum boundary value of the dominant direction is set to 8, the Ψ value is greater than the boundary, and it is judged as a significant offset type. This numerical result is used in the construction of the directional classification set to guide the path to be classified as a non-dominant offset direction type trajectory. The angle is expressed in degrees, the distance is expressed in meters, and the number and sector number are dimensionless quantities. The innovation of this formula lies in the introduction of the spatial path length structure (S1+S2), the path number correction (μ+ε), and the combined comparison relationship between the crossing distribution η and the angular average δ, so that the trajectory offset direction classification no longer depends solely on the angle difference, but is judged by integrating structural and spatial continuity. The result shows that the current trajectory has a significant directional offset, and its classification result should be attributed to the non-dominant offset type, and is used as the particle offset direction classification result for subsequent judgment and identification records.

[0098] The offset identification recording submodule calls the particle offset direction classification results, filters the spatial identification points with inconsistent directional labels, matches the corresponding path trajectory numbers and the monitoring time periods to which they belong, marks the spatial coordinates and sequence numbers of the points, and obtains the particle path offset identification.

[0099] The coordinates of the monitoring points associated with the end points of each path are indexed and bound to their directional classification labels. Then, all spatial coordinate points are traversed to screen out points that are located in the same sector in the same time period but have obvious differences in directional labels. Such points mark sudden changes, turning points, or inconsistent directional projections of the airflow path. During the screening process, it is determined whether the label difference reaches the recognition threshold. The difference is calculated as a discrete value between the directional codes. If it exceeds three levels of distinction, it is considered a difference point. The spatial coordinates of the difference point and its corresponding time series number are extracted and uniformly recorded in the path deviation identification sequence table. At the same time, the trajectory numbers of all recorded points are attached to the record items to support subsequent association analysis operations. Finally, a complete data set is integrated to obtain the particle path deviation identifier.

[0100] See also Figure 5 , the boundary recognition and analysis module includes:

[0101] The sampling optical monitoring submodule collects the incident path data of the light beam in the visible light and near-infrared bands based on the sampling points located in the particle path offset mark, monitors the initial values ​​and time series changes of the incident angle and refraction angle at the sampling points, and obtains the incident refraction angle change group.

[0102] By adjusting the emission direction of the light source and the position of the optical path channel, the incident and outgoing beam paths under multiple angle conditions are obtained, and the refraction angle change value is further collected. In the specific implementation process, the spindle area and the tool change position in the CNC equipment are used as the key sampling areas, and optical sampling points are arranged in them. The incident angle of the light source is set to 30°, 45°, and 60°, and the beam refraction angle value under each gear is recorded. Combined with the incident medium being a mixed environment of oil mist and air, the optical path difference and the change in the wavelength extension direction are compared, and the initial value of the refraction angle is obtained according to the trigonometric function relationship. The measurement system is used to record the change of the angle of the refraction trajectory under the same path with the time series. For example, under 45° incidence, the first sampling refraction angle is 28°, and the subsequent recorded refraction angle gradually increases to 33° with the change of particle concentration. This indicates that the particle boundary moves within the path. Multiple measurement values ​​within the path are used to establish the incidence and refraction angle change curve, and then obtain its change value group. The above process is repeated for the sampling path, and the change groups under multiple paths are grouped and stored for subsequent calculation, and finally the incident refraction angle change group is obtained.

[0103] The refraction change calculation submodule extracts the deflection trajectory and particle width change value of the light beam penetrating the particle path according to the incident refraction angle change group, and performs cross-time series weighted accumulation using the formula:

[0104] The boundary bending strength value is calculated to obtain the boundary recognition interference index.

[0105] Among them, Δθ represents the boundary bending strength value, Q a represents the change in the incident angle of visible light at the ath sampling position, β a represents the change in near-infrared refraction angle at the ath sampling position, F a represents the deflection length of the penetration trajectory at the ath sampling position, γ a represents the particle size path offset width at the ath sampling position, λ a represents the beam propagation timing number of the ath sampling position, and z is the total number of sampling positions.

[0106] The incident angle and refraction angle data of each sampling point in the visible light and near-infrared bands are obtained from the monitoring, and are processed in combination with the trajectory deflection length and particle size path width change value in the path of the light beam penetrating the particles. The optical monitoring equipment is deployed at the boundary of the CNC processing area, and the sampling path is set in the high-density particle distribution section. Each path has multiple sets of optical characteristic value changes in different time series. During the acquisition process, Q a It represents the incident angle change value of the visible light at the sampling point a, in degrees, which is calculated by the laser scanning system through the angle difference. For example, when the initial incident angle is 32° and the refraction angle is 28°, then Q a =4,β a represents the refraction angle change value of the near-infrared band at position a. Assuming that the incident angle at this point is 35° and the refraction angle is 29°, then β a =6. a It represents the deflection length of the beam penetration segment in the particle path at position a, in millimeters (mm). The light path change trajectory is recorded and integrated by the three-dimensional tracking module. Assume that the deflection path is 6.5 mm. a It indicates the particle size path offset width of the corresponding segment, and the unit is also in millimeters. It is obtained by multiplying the particle concentration distribution density by the average particle size radius. Assume that the particle size path width of this segment is 2.5mm. a is the time series number of the beam propagation at the sampling point, corresponding to the sampling order, for example, the third frame is λ a =3.

[0107] Assume three sets of data:

[0108] Group 1: Q1=4, β1=6, F1=6.5, γ1=2.5, λ1=1;

[0109] Group 2: Q2 = 5, β2 = 7, F2 = 7.2, γ2 = 3.0, λ2 = 2;

[0110] Group 3: Q3=3, β3=4, F3=5.0, γ3=2.0, λ3=3;

[0111] The calculation process is as follows:

[0112] Item 1:

[0113] Item 2:

[0114] Item 3:

[0115] Adding the three terms together, we get: Δθ = 20 + 49.176 + 36.372 = 105.548;

[0116] The boundary bending strength value is 105.548. By constructing a fusion index together with the angle change, spatial path change and time series, a fusion judgment of cross-dimensional and multi-source data is formed, which is conducive to accurately evaluating the dynamic boundary fluctuations in the particle interference area. The results show that the intensity of boundary refraction in the current particle path has a fluctuation range at multiple time points, the overall refraction deflection trend is active, and the boundary identification interference index value is high. It is necessary to screen and process spatial clustering anomalies in subsequent modules.

[0117] The abnormal boundary aggregation submodule calls the boundary identification interference index, extracts the spatial distribution of nodes whose values ​​are higher than the interference threshold, aggregates the label boundary data by value and position, and obtains the boundary identification abnormal distribution data.

[0118] The identified abnormal boundary data is extracted and aggregated into labels based on their spatial positions. This provides an efficient tool for quality control in the manufacturing process. Through real-time analysis of beam refraction data, non-uniform areas of coating thickness can be quickly identified, and timely adjustments can be made to ensure the uniformity and quality of the coating. Nodes with values ​​exceeding the limit are identified through preset thresholds, and then cluster analysis is performed based on the spatial positions of the nodes to construct a spatial distribution map. This map helps operators quickly locate problem areas and take corresponding measures, such as adjusting spraying parameters or correcting operation paths, to eliminate or reduce defects and ensure that the product meets design specifications. This method not only improves production efficiency, but also reduces material waste and rework rates, and obtains boundary recognition abnormal distribution data.

[0119] See also Figure 6 , the separation efficiency feedback module includes:

[0120] The boundary point analysis submodule identifies abnormal distribution data based on the boundary, identifies distribution points that meet the conditions, and determines whether each distribution point meets the set boundary conditions according to its spatial position. It calibrates the distribution points that meet the conditions and generates distribution point boundary information.

[0121] The spatial position data of all distribution points in the fan adjustment area are collected through sensors or real-time monitoring. The data is preliminarily screened to remove points that deviate from the normal distribution to ensure the validity of the data. For each distribution point, its specific position in the adjustment area is located through coordinate transformation or spatial mapping methods, and then through predetermined boundary conditions. For example, based on the radius of the maximum response area of ​​the fan or a specific geometric model, it is determined whether the point meets the boundary conditions. If the conditions are met, the point is identified as a valid boundary point. The applied boundary conditions can be static set values ​​or dynamic set values ​​based on environmental changes. For example, in the response area affected by the change in fan speed, if the fan adjustment area is circular and the radius is set to 10 meters, the distribution points less than 10 meters away from the center point are valid boundary points. All valid distribution points are organized into a data set to generate distribution point boundary information.

[0122] The fan adjustment state identification submodule analyzes the coverage quantity of the distribution points in the fan adjustment area based on the distribution point boundary information, and compares it with the fan response state to obtain the fan response adjustment state data.

[0123] First, calculate the coverage area of ​​each distribution point in the fan adjustment area. Determine the relative position of each distribution point according to the movement trajectory and adjustment range of the fan. By analyzing the data one by one, the number of distribution points in each area is obtained, and then the response adjustment state of the fan is judged. The judgment standard of the response state is based on the relationship between the density of distribution points in the area and the wind speed or flow generated by the fan adjustment. If the number of distribution points exceeds a certain threshold, the fan response state of the area is considered to be "high response", otherwise it is "low response". In this process, the response state threshold set depends on the expected performance of the fan and the coverage range of the airflow in the area. If the number of distribution points in the fan adjustment area reaches or exceeds the predetermined threshold of 20, the area is judged to be in a "high response" state, and the fan response adjustment state data is obtained.

[0124] The regional stability calibration submodule analyzes the overlapping relationship between the fan adjustment area and the distribution point density according to the fan response adjustment status data, calibrates the overlapping area number, and outputs the monitoring area separation stability distribution.

[0125] By counting the number of distribution points and the fan response status in each area, the distribution of response points in each area is calculated, and it is determined whether there is an overlapping area. The judgment standard of the overlapping area is based on the relationship between the number of distribution points in the area and the fan response range. If the number of distribution points in a certain area coincides with the fan response area and meets the set threshold conditions, the area is marked as an overlapping area. All overlapping areas are marked by numbering for subsequent analysis and adjustment. If the number of distribution points in a certain area is 15, and the overlap between the response area it covers and the maximum response area of ​​the fan is greater than 80%, the area is considered to be an overlapping area and is numbered, and the monitoring area separation stability distribution is output.

[0126] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. Machining oil mist separation real-time monitoring system, characterized by: The system comprises: The particle capture monitoring module collects the oil mist particle concentration and air flow velocity of the CNC spindle exhaust channel and tool change position based on the machining oil mist separation monitoring data, analyzes the change of particle capture efficiency at the collection point, determines the offset from the concentration reference value, identifies the offset direction, and generates a particle concentration trend label at the monitoring position; an airflow adjustment control module, based on the particle concentration trend tag of the monitoring position, identifying the fan orientation of the concentration rising area, determining whether the airflow adjustment response is lower than the fan linkage adjustment reference value, marking the position that does not meet the response standard, and outputting a fan response status list; The particle path judgment module calls the fan response state list, identifies the particle settling direction trajectory in the airflow path, performs direction classification on the particle offset direction, records the identification information of the differentiated directional offset point, and obtains the particle path offset identification; The boundary recognition and analysis module calls the sampling points located in the particle path offset mark, collects the visible light and near-infrared refraction angles under the incident path of the light beam, analyzes the changes in the incident and refraction angles when the light beam penetrates the particles, classifies and stores the boundary feature anomaly points, and obtains the boundary recognition anomaly distribution data.

2. The machining oil mist separation real-time monitoring system according to claim 1 is characterized in that: The particle concentration trend label of the monitoring position includes the concentration change amplitude, change trend type, and abnormal fluctuation frequency; the fan response status list includes response delay parameters, guide angle offset, and wind speed adjustment amplitude; the particle path offset identifier includes particle sedimentation direction, offset path characteristics, and directional classification number; the boundary identification abnormal distribution data includes optical refraction difference, abnormal boundary density, and refraction abnormal distribution range.

3. The machining oil mist separation real-time monitoring system according to claim 1 is characterized in that: The particle capture monitoring module comprises: The particle concentration monitoring submodule extracts the air flow velocity, particle concentration, channel cross-sectional area and sampling time of the CNC spindle exhaust channel and tool change position based on the machining oil mist separation monitoring data, identifies the particle throughput per unit time, analyzes the particle distribution differences within the same time period, and obtains the particle transfer volume change range; The offset identification submodule compares the difference between the particle transfer amount change interval and the particle concentration change reference value, and combines the particle concentration change speed of the monitoring point with the offset rate mark of the transfer interval, using the formula: Calculate the particle concentration offset intensity value and obtain the concentration offset characteristic value; Where ΔC represents the particle concentration offset intensity value, v i represents the airflow velocity at the i-th monitoring point, q i represents the particle concentration at the i-th monitoring point, B i Represents the baseline value of particle concentration change at the i-th monitoring point, A i represents the channel cross-sectional area of ​​the i-th monitoring point, t i represents the sampling time series index of the i-th monitoring point, and n represents the total number of monitoring points; The trend label identification submodule calls the concentration offset feature, determines the offset progressive direction and accumulation, matches the trend extreme point coordinates, and generates a particle concentration trend label at the monitoring location.

4. The machining oil mist separation real-time monitoring system according to claim 3 is characterized in that: The airflow adjustment control module comprises: The wind direction labeling submodule identifies the area with increasing concentration based on the particle concentration trend label of the monitoring position, selects the fan numbers with large steering angles and low wind speeds, and generates a fan steering label set; The response judgment submodule calls the number in the fan guide tag set to identify the corresponding real-time guide angle and wind speed value, using the formula: Calculate the airflow response offset value, filter the numbers whose offset exceeds the limit, and obtain the number set of non-compliant responses; Wherein, R represents the airflow response offset value, α represents the real-time steering angle, D represents the set steering angle reference value, w represents the real-time wind speed value, L represents the distance between the steering point and the fan outlet, τ represents the response monitoring delay sequence number, and θ represents the fan adjustment reference angle difference; The status output submodule combines the fan numbers and corresponding response statuses according to the response non-standard number set, sets the response status labels, enters all the corresponding relationships between numbers and statuses into the fan status array, and outputs the fan response status list.

5. The machining oil mist separation real-time monitoring system according to claim 4 is characterized in that: The particle path judgment module includes: The path trajectory extraction submodule calls the guide angle and the spatial position coordinates of the monitoring point in the fan response state list, selects the sectors whose projection difference exceeds the set threshold, analyzes the degree of coincidence between the particle concentration distribution and the wind direction in the path section, and obtains the sedimentation direction trajectory sequence; The directional classification submodule extracts the offset angle, monitoring point coordinate set and trajectory boundary parameters based on the settlement direction trajectory sequence, and classifies the trajectory according to the wind direction reference angle and spatial distribution, using the formula: Calculate the particle direction offset value, make direction attribution judgment, and generate the particle offset direction classification result; Among them, Ψ represents the particle direction offset value, represents the trajectory deviation angle, E represents the dominant wind direction angle of the area, S1 represents the horizontal projection distance from the starting point of the trajectory to the monitoring point, S2 represents the vertical projection distance from the end point of the trajectory to the monitoring point, μ represents the trajectory number sequence value, ε represents the projection layer offset, η represents the number of sectors crossed by the trajectory, and δ represents the average trajectory deviation angle; The offset identification recording submodule calls the particle offset direction classification result, filters the spatial identification points with inconsistent directional labels, matches the corresponding path trajectory number and the monitoring time period, marks the spatial coordinates and sequence number of the point, and obtains the particle path offset identification.

6. The machining oil mist separation real-time monitoring system according to claim 5 is characterized in that: The boundary recognition and analysis module includes: The sampling optical monitoring submodule collects incident path data of the light beam in the visible light and near infrared bands based on the sampling points located in the particle path deviation mark, monitors the initial values ​​and time series changes of the incident angle and refraction angle at the sampling points, and obtains a group of incident and refraction angle changes; The refraction change calculation submodule extracts the deflection trajectory and particle width change value in the path of the light beam penetrating the particle according to the incident refraction angle change group, and performs cross-time series weighted accumulation using the formula: Calculate the boundary bending strength value to obtain the boundary identification interference index; Among them, Δθ represents the boundary bending strength value, Q a represents the change in the incident angle of visible light at the ath sampling position, β a represents the change in near-infrared refraction angle at the ath sampling position, F a represents the deflection length of the penetration trajectory at the ath sampling position, γ a represents the particle size path offset width at the ath sampling position, λ a represents the beam propagation timing number of the ath sampling position, and z is the total number of sampling positions; The abnormal boundary aggregation submodule calls the boundary identification interference index, extracts the spatial distribution of nodes whose values ​​are higher than the interference threshold, aggregates the label boundary data according to the value and position, and obtains the boundary identification abnormal distribution data.

7. The machining oil mist separation real-time monitoring system according to claim 1 is characterized in that: The system also includes a separation efficiency feedback module: A separation efficiency feedback module determines whether there is an overlapping relationship between the distribution density and the fan response state based on the coverage number of the distribution points in the boundary recognition abnormal distribution data within the fan adjustment area, marks the number of the overlapping area, and outputs the separation stability distribution of the monitoring area; The monitoring area separation stability distribution includes overlapping area identification, stability level classification, and density coverage ratio.

8. The machining oil mist separation real-time monitoring system according to claim 7, characterized in that: The separation efficiency feedback module comprises: The boundary point analysis submodule identifies abnormal distribution data based on the boundary, identifies distribution points that meet the conditions, and determines whether the set boundary conditions are met according to the spatial position of each distribution point, calibrates the distribution points that meet the conditions, and generates distribution point boundary information; A fan adjustment state identification submodule analyzes the coverage quantity of the distribution points in the fan adjustment area based on the distribution point boundary information, and compares it with the fan response state to obtain fan response adjustment state data; The regional stability calibration submodule analyzes the overlapping relationship between the fan adjustment area and the distribution point density according to the fan response adjustment state data, calibrates the overlapping area number, and outputs the monitoring area separation stability distribution.

Images