Detection data intelligent management system and method applied to reaction kettle stirrer
By installing an annular fixture and infrared sensors inside the reactor agitator, initial distance sequences are collected and processed to calculate runout deviation and regional severity index. The inspection mode is dynamically adjusted, which solves the problems of uneven mixing and equipment wear caused by radial runout of the agitator, and achieves efficient and accurate fault monitoring and early warning.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- NANJING SANYE AGITATOR SYST CO LTD
- Filing Date
- 2026-02-03
- Publication Date
- 2026-05-05
AI Technical Summary
In the prior art, radial runout of the reactor agitator leads to unstable contact between the agitator blades and the reactor wall or materials, resulting in uneven material mixing, increased wear of components, increased equipment vibration and noise, and the sensor layout is not flexible enough, which may cause the most severe runout area to be missed.
An annular fixture is installed inside the reactor agitator, and an infrared sensor is deployed. The initial distance sequence is collected by the infrared sensor, and the baseline distance sequence and center coordinates are obtained by processing with a mean filtering algorithm. The instantaneous jump deviation is calculated, and the jump deviation threshold and regional severity index are preset. The inspection mode is dynamically adjusted to achieve accurate monitoring and graded early warning of potential abnormal areas.
It enables efficient and accurate monitoring of the agitator, timely detection of potential risks, prevention of escalation of faults, ensuring continuous and safe operation of the equipment, reducing resource consumption, and improving monitoring efficiency and accuracy.
Smart Images
Figure CN121979063A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of data analysis technology, specifically to an intelligent management system and method for detection data applied to a reactor agitator. Background Technology
[0002] Reactors are core reaction equipment in the process industry. As a key moving component, the agitator operates under harsh conditions such as high temperature, high pressure, corrosion, flammability, and explosion. Its operating status directly determines production safety, product quality, and energy consumption. Meanwhile, the data management technology for reactor agitator testing has gradually evolved from manual recording and offline testing to intelligent management and closed-loop operation and maintenance, accompanying the automation, informatization, and intelligentization of the process industry.
[0003] When the agitator of a reactor has large radial runout, it will cause unstable contact between the agitator blades and the reactor wall or materials during rotation, resulting in uneven material flow and mixing. At the same time, radial runout will also subject components such as the agitator shaft, bearings, and seals to additional stress and friction, leading to accelerated wear of these components, increased motor load, and motor overheating. Finally, radial runout will cause increased vibration and noise in the equipment, which will not only affect the operating environment but may also cause other components to loosen or be damaged. Furthermore, the traditional fixed layout of multi-sensor sampling points is fixed and not flexible enough, which may lead to missed detection of the areas with the most severe runout. Summary of the Invention
[0004] The purpose of this invention is to provide an intelligent management system and method for detection data applied to reactor agitators, in order to solve the problems raised in the prior art.
[0005] To achieve the above objectives, the present invention provides the following technical solution:
[0006] A method for intelligent management of detection data applied to a reactor agitator, the method comprising the following steps:
[0007] S100. Set up an annular fixture inside the stirrer of the reactor, deploy an infrared sensor on the annular fixture, use the infrared sensor to collect the initial distance sequence when the stirrer is working, process the initial distance sequence to obtain the reference distance sequence, reference center coordinates and reference circle radius.
[0008] Furthermore, the specific steps for processing the initial distance sequence to obtain the reference distance sequence, reference center coordinates, and reference circle radius are as follows:
[0009] S101. After the reactor agitator is started and reaches its rated speed, an infrared sensor is used to rotate uniformly around the annular fixture for one revolution, collecting initial distance sequences D0={d1, d2, ..., dn} at fixed angle intervals of θ. n}, d1, d2, ..., dn The initial distances represent the distances at the 1st, 2nd, ..., nth positions acquired by the infrared sensor, where each initial distance represents the distance between the infrared sensor and the stirring shaft of the stirrer. A baseline distance sequence D is obtained by processing the initial distance sequence using a mean filtering algorithm. base The initial distance sequence is collected by uniformly rotating an infrared sensor, and then processed by a mean filtering algorithm to obtain a baseline distance sequence. This effectively filters out random noise in the initial data and ensures the stability of the baseline data. The infrared sensor acquisition interval angle is limited to θ ≤ 1° to ensure that the collected initial distance sequence covers the entire circumference of the stirring shaft, and the sampling density is sufficient so that the fitted baseline parameters can truly reflect the shaft center position and radius dimensions under normal operating conditions.
[0010] S102, Utilizing the inner radius R of the ring-shaped tooling ring The distance value measured by the infrared sensor is converted into polar coordinates of a point on the surface of the stirring shaft, using the formula: ρ i =R ring -d i , where ρ i d represents the radial distance from the center of the annular tool at the i-th position to the surface of the stirring shaft. i This represents the baseline distance at the i-th position in the baseline distance sequence;
[0011] By collecting the angle at each location and combining it with the corresponding radial distance, the Cartesian coordinates (x, y, z) of the point on the surface of the stirring shaft at each location are calculated. i y i );
[0012] Let the equation of the circle to be fitted be (x-x0). 2 +(y-y0) 2 =R 2 Let (x0, y0) represent the coordinates of the reference center to be determined, and R represent the radius of the reference circle to be determined; let A = 2x0, B = 2y0, and C = R. 2 -x0 2 -y0 2 Then the equation of the circle is transformed into a linear equation: x 2 +y 2 =Ax+By+C; Substitute the Cartesian coordinates of all stirring shaft surface points transformed from the baseline distance sequence into (x, y) of the linear equation, and use the least squares solution to calculate A, B, and C;
[0013] Substituting A, B, and C into A=2x0, B=2y0, and C=R respectively... 2 -x0 2 -y0 2The coordinates of the reference center (x0, y0) and the radius R of the reference circle are calculated. By transforming polar coordinates to Cartesian coordinates and fitting the circle equation using the least squares solution, the coordinates of the reference center and the radius of the reference circle are accurately calculated. This provides an objective and unified reference standard for subsequent judgment of runout deviation, avoiding errors caused by subjectively setting the reference.
[0014] S200: When the agitator is working, set the monitoring cycle and uniform speed inspection mode. The infrared sensor collects the real-time distance value according to the monitoring cycle, and calculates the instantaneous jump deviation using the real-time distance value and the radius of the ring tool.
[0015] Furthermore, the specific steps for calculating the radial offset and instantaneous runout deviation using the real-time distance value and the radius of the annular tooling are as follows:
[0016] S201. Set the monitoring cycle T. The uniform speed inspection mode means that the infrared sensor moves at a fixed angle S. θ For interval sampling, a uniformly rotating VP performs a patrol inspection, collecting the distance at each location and treating each location as a sampling point; an infrared sensor collects the real-time distance d at each sampling point. current The instantaneous runout deviation of the current sampling point corresponding to the stirring shaft surface relative to the reference center is calculated using the inner radius of the annular fixture and the radius of the reference circle. The formula is: E=|(R ring -d current E represents the instantaneous fluctuation deviation. By setting the monitoring period T and the uniform speed inspection mode, the infrared sensor collects real-time distance values at fixed angular intervals and with uniform rotation, enabling continuous monitoring of the agitator's operating status and timely capture of instantaneous position changes of the agitator shaft. Parameters such as the monitoring period T and the acquisition angle interval can be flexibly adjusted to adapt to different monitoring scenarios based on the operating characteristics of the reactor agitator and actual monitoring needs, thus improving the versatility of the solution.
[0017] S300: Preset jump deviation threshold, use the jump deviation threshold to judge the instantaneous jump deviation to obtain potential abnormal areas, calculate the area severity index of potential abnormal areas, and use the area severity index to set the monitoring priority of all potential abnormal areas;
[0018] Furthermore, the specific steps for setting the monitoring priority of all potentially abnormal areas using the regional severity index are as follows:
[0019] S301, Preset fluctuation deviation threshold E threshold Within a sampling period, when the instantaneous fluctuation deviation E of consecutive sampling points > fluctuation deviation threshold E thresholdWhen a continuous angular region is marked as a "potential abnormal region", the number of continuous sampling points is greater than 1; a preset jump deviation threshold is used to mark potential abnormal regions by judging whether the instantaneous jump deviation of continuous sampling points exceeds the standard, which avoids misjudgment caused by the error of a single sampling point and improves the accuracy of anomaly identification.
[0020] S302. Calculate the regional severity index for each potential abnormal region identified by the marking. The formula is as follows:
[0021] ;
[0022] In the formula, S represents the regional severity index, and E' represents the average instantaneous jump deviation of all sampling points within the potentially abnormal region. The instantaneous jump deviation standard deviation represents all sampling points, L represents the radians corresponding to the angle of the potential anomaly area, and w1, w2, and w2 represent the weights of the average instantaneous jump deviation, the instantaneous jump deviation standard deviation, and the radians, respectively, set by the staff, with w1+w2+w2=1; a regional severity index S is introduced to comprehensively consider the average instantaneous jump deviation, the deviation standard deviation, and the angle radians of the potential anomaly area, fully reflecting the severity of the anomaly area and avoiding the one-sidedness caused by judging based on only a single indicator.
[0023] All potentially abnormal areas are sorted from highest to lowest severity index, and monitoring priorities are assigned based on this ranking: the area ranked first has the highest monitoring priority, and the priority decreases sequentially from the first ranked area. Prioritizing potentially abnormal areas based on their severity index ensures that monitoring resources are concentrated on areas with high severity and risk, improving monitoring efficiency and avoiding resource waste.
[0024] S400: When a high monitoring priority is detected, the uniform speed inspection mode is interrupted, and intensive scanning of key areas is performed.
[0025] Furthermore, the specific steps for performing intensive scanning of key areas are as follows:
[0026] S401, Preset priority threshold P threshold If a potential anomaly is identified as having high priority, and the monitoring priority of the area is greater than or equal to the priority threshold, the uniform speed inspection mode is interrupted, and a dense scan of the key area is performed. Specifically, the dense scan of the key area involves setting the dense angle α and the key scan period T. important Constraint α θ T important >T; When the monitoring priority of a potential abnormal area is less than the priority threshold, the uniform speed inspection mode continues; when the monitoring priority of a potential abnormal area reaches the threshold, the uniform speed inspection mode is interrupted, and intensive scanning of key areas is started. By reducing the acquisition angle interval α and shortening the key scanning cycle, refined monitoring of abnormal areas is achieved, and more comprehensive abnormal data is obtained.
[0027] S402. If a potential anomaly area spreads or moves during a continuous monitoring cycle, the infrared sensor is used to predict the spread and movement area of the potential anomaly area in the next monitoring cycle based on the direction of spread and movement within the continuous monitoring cycle, and to conduct advance tracking and monitoring. Predicting and tracking the spread and movement area in the next monitoring cycle based on the spread and movement of potential anomalies allows for early detection of anomaly expansion trends, buying time for subsequent fault handling. Intensive scanning is performed only on high-priority anomaly areas, while low-priority areas continue to be inspected at a constant speed. This ensures monitoring accuracy in key areas while avoiding the decrease in monitoring efficiency and increased resource consumption caused by intensive scanning of the entire area.
[0028] S500: Within a key scanning cycle, integrate all sampling points collected by the infrared sensor, use interpolation to reconstruct the approximate axis trajectory of the stirring shaft relative to the reference center coordinates, and calculate the comprehensive runout index within the key scanning cycle.
[0029] Furthermore, the specific steps for calculating the comprehensive fluctuation index within the monitoring period are as follows:
[0030] S501. Store all sampling points of the potential abnormal area in the key area intensive scanning within a key scanning cycle, and use the interpolation algorithm to reconstruct the approximate axis trajectory of the stirring shaft relative to the reference center within the key scanning cycle based on all discrete sampling points; use the interpolation algorithm to reconstruct the approximate axis trajectory based on discrete sampling points, and transform the abstract jumping data into an intuitive trajectory graph, so that the staff can quickly understand the offset pattern and abnormal characteristics of the stirring shaft.
[0031] S502. Calculate the comprehensive runout index within the key scanning period. The comprehensive runout index includes the maximum runout, the average runout, the main runout direction, and the trajectory ellipticity. The maximum runout represents the maximum value of the instantaneous runout deviation of all sampling points within the key scanning period. The average runout represents the average value of the instantaneous runout deviation of all sampling points within the key scanning period.
[0032] The main direction of the jump is as follows: within the key scanning period, the absolute value of the instantaneous jump deviation of all sampling points is removed as the offset vector. The offset vector includes the offset amount and the direction, where the direction represents the direction of the unbalance center during the jump from the reference center. The sum of all offset vectors is calculated, and the direction of the sum of the offset vectors is taken as the main direction of the jump.
[0033] The ellipticity of the trajectory is obtained by calculating the ratio of the major and minor axes of the approximate axis-centered trajectory.
[0034] By calculating comprehensive runout indicators such as maximum runout, average runout, main runout direction, and trajectory ellipticity, the runout of the stirring shaft is reflected from multiple dimensions, which can better reveal the operating status of the stirrer compared to a single indicator.
[0035] S600: Preset dynamic threshold library; use dynamic threshold library to judge comprehensive fluctuation index and perform graded early warning; set fault judgment rules based on historical agitator faults; analyze and judge real-time fault type based on approximate shaft center trajectory.
[0036] Furthermore, the specific steps for comparing the approximate axis trajectory with faults in the fault type database to determine the real-time fault type are as follows:
[0037] S601. The preset dynamic threshold library includes warning thresholds, alarm thresholds, and shutdown thresholds. The dynamic threshold library is used to determine the comprehensive fluctuation index and then issue graded warnings. Specifically:
[0038] When the average fluctuation exceeds the warning threshold, a Level 1 warning is triggered, and a prompt is displayed on the operation interface, while the device continues to operate.
[0039] When the maximum fluctuation exceeds the alarm threshold, a level two warning is triggered, and an audible and visual alarm is activated.
[0040] When the maximum fluctuation exceeds the shutdown threshold, a Level 3 emergency shutdown command is triggered. The system automatically interlocks, stops the mixer, and locks the equipment to prevent accidental restart. Simultaneously, an alarm is sent to staff. A preset dynamic threshold library triggers different levels of warnings based on comprehensive fluctuation indicators. This avoids resource waste caused by excessive warnings for minor anomalies and safety accidents caused by insufficient warnings for serious anomalies. In the Level 3 warning mechanism, the Level 3 emergency shutdown command automatically interlocks and locks the equipment when a serious malfunction occurs, effectively preventing the fault from escalating and avoiding equipment damage and personnel safety risks. Simultaneously, alarm information is sent to staff to ensure timely handling of the fault.
[0041] S602. Preset ellipse threshold and set fault judgment rules based on historical stirrer faults, specifically:
[0042] When the ellipticity of the approximate axis trajectory is less than the elliptic threshold and the main direction of the jump is less than ±10 degrees in the continuous key scanning cycle, the fault is judged as "static quality imbalance".
[0043] When the ellipticity of the approximate axis trajectory is greater than or equal to the elliptic threshold and the main direction of the jump changes periodically within the continuous key scanning cycle, the fault is judged as "axis bending or dynamic misalignment".
[0044] When the approximate axis trajectory is irregular, the fault is judged as "loose bearing or excessive clearance".
[0045] The intelligent management system for detection data applied to reactor agitators includes a baseline calculation module, a real-time fluctuation module, a fault area identification module, an inspection mode adjustment module, a comprehensive analysis module, and a fault early warning module.
[0046] The reference calculation module is used to set up an annular fixture inside the reactor stirrer, deploy an infrared sensor on the annular fixture, use the infrared sensor to collect the initial distance sequence when the stirrer is working, and process the initial distance sequence to obtain the reference distance sequence, reference center coordinates and reference circle radius;
[0047] The real-time jumping module is used to set the monitoring cycle and uniform speed inspection mode when the stirrer is working. The infrared sensor collects the real-time distance value according to the monitoring cycle, and calculates the instantaneous jumping deviation using the real-time distance value and the radius of the annular tool.
[0048] The fault area identification module is used to preset the fluctuation deviation threshold, use the fluctuation deviation threshold to judge the instantaneous fluctuation deviation to obtain potential abnormal areas, calculate the area severity index of the potential abnormal areas, and use the area severity index to set the monitoring priority of all potential abnormal areas.
[0049] The inspection mode adjustment module is used to interrupt the uniform speed inspection mode and perform intensive scanning of key areas when a high monitoring priority is detected.
[0050] The comprehensive analysis module is used to integrate all sampling points collected by the infrared sensor within a monitoring cycle, reconstruct the approximate axis trajectory of the stirrer's stirring shaft relative to the reference center coordinates using interpolation, and calculate the comprehensive runout index within the monitoring cycle.
[0051] The fault early warning module is used to preset a dynamic threshold library, and after judging the comprehensive fluctuation index using the dynamic threshold library, it performs graded early warning. It sets fault judgment rules based on historical agitator faults and analyzes the approximate shaft center trajectory to judge the real-time fault type.
[0052] The comprehensive analysis module includes a center trajectory unit and a comprehensive bounce index unit;
[0053] The axis center trajectory unit is used to reconstruct the approximate axis center trajectory of the stirring shaft relative to the reference center within the key scanning period based on all discrete sampling points using an interpolation algorithm.
[0054] The comprehensive bounce index unit is used to calculate the maximum bounce, average bounce, main bounce direction, and trajectory ellipticity.
[0055] The fault early warning module includes an early warning unit and a fault type unit;
[0056] The early warning unit is used to preset a dynamic threshold library including early warning thresholds, alarm thresholds and shutdown thresholds, and to perform graded early warnings after judging the comprehensive fluctuation index using the dynamic threshold library.
[0057] The fault type unit is used to preset the ellipse threshold and set fault judgment rules based on historical stirrer faults.
[0058] Compared with the prior art, the beneficial effects of the present invention are:
[0059] 1. This invention achieves a differentiated monitoring strategy of "precise monitoring of key areas and efficient inspection of ordinary areas" by prioritizing potential abnormal areas and dynamically adjusting the inspection mode. This ensures monitoring quality while reducing unnecessary resource consumption and improving overall monitoring efficiency.
[0060] 2. The combination of real-time monitoring and hierarchical early warning mechanisms can promptly detect potential risks in the early stages of anomalies and even predict the spread of anomalies in advance, allowing staff sufficient time to handle faults, effectively preventing the escalation of faults, and ensuring the continuous and safe operation of the reactor agitator. Attached Figure Description
[0061] Figure 1 This is a schematic diagram illustrating the steps of the intelligent management method for detection data applied to a reactor agitator according to the present invention. Detailed Implementation
[0062] 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 embodiments of the present invention, and not all embodiments. 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.
[0063] Example: Figure 1 As shown, the present invention provides a technical solution.
[0064] A method for intelligent management of detection data applied to a reactor agitator, the method comprising the following steps:
[0065] S100. Set up an annular fixture inside the stirrer of the reactor, deploy an infrared sensor on the annular fixture, use the infrared sensor to collect the initial distance sequence when the stirrer is working, process the initial distance sequence to obtain the reference distance sequence, reference center coordinates and reference circle radius.
[0066] The specific steps for processing the initial distance sequence to obtain the reference distance sequence, reference center coordinates, and reference circle radius are as follows:
[0067] S101. After the reactor agitator is started and reaches its rated speed, an infrared sensor is used to rotate uniformly around the annular fixture for one revolution, collecting initial distance sequences D0={d1, d2, ..., dn} at fixed angle intervals of θ. n}, d1, d2, ..., d n The initial distances represent the distances at the 1st, 2nd, ..., nth positions acquired by the infrared sensor, where each initial distance represents the distance between the infrared sensor and the stirring shaft of the stirrer. A baseline distance sequence D is obtained by processing the initial distance sequence using a mean filtering algorithm. base The initial distance sequence is collected by uniformly rotating an infrared sensor, and then processed by a mean filtering algorithm to obtain a baseline distance sequence. This effectively filters out random noise in the initial data and ensures the stability of the baseline data. The infrared sensor acquisition interval angle is limited to θ ≤ 1° to ensure that the collected initial distance sequence covers the entire circumference of the stirring shaft, and the sampling density is sufficient so that the fitted baseline parameters can truly reflect the shaft center position and radius dimensions under normal operating conditions.
[0068] S102, Utilizing the inner radius R of the ring-shaped tooling ring The distance value measured by the infrared sensor is converted into polar coordinates of a point on the surface of the stirring shaft, using the formula: ρ i =R ring -d i , where ρ i d represents the radial distance from the center of the annular tool at the i-th position to the surface of the stirring shaft. i This represents the baseline distance at the i-th position in the baseline distance sequence;
[0069] By collecting the angle at each location and combining it with the corresponding radial distance, the Cartesian coordinates (x, y, z) of the point on the surface of the stirring shaft at each location are calculated. i y i );
[0070] Let the equation of the circle to be fitted be (x-x0). 2 +(y-y0) 2 =R 2 Let (x0, y0) represent the coordinates of the reference center to be determined, and R represent the radius of the reference circle to be determined; let A = 2x0, B = 2y0, and C = R. 2 -x0 2 -y0 2 Then the equation of the circle is transformed into a linear equation: x 2 +y 2 =Ax+By+C; Substitute the Cartesian coordinates of all stirring shaft surface points transformed from the baseline distance sequence into (x, y) of the linear equation, and use the least squares solution to calculate A, B, and C;
[0071] Substituting A, B, and C into A=2x0, B=2y0, and C=R respectively... 2 -x0 2 -y0 2 The coordinates of the reference center (x0, y0) and the radius R of the reference circle are calculated. By transforming polar coordinates to Cartesian coordinates and fitting the circle equation using the least squares solution, the coordinates of the reference center and the radius of the reference circle are accurately calculated. This provides an objective and unified reference standard for subsequent judgment of runout deviation, avoiding errors caused by subjectively setting the reference.
[0072] S200: When the agitator is working, set the monitoring cycle and uniform speed inspection mode. The infrared sensor collects the real-time distance value according to the monitoring cycle, and calculates the instantaneous jump deviation using the real-time distance value and the radius of the ring tool.
[0073] The specific steps for calculating the radial offset and instantaneous runout deviation using the real-time distance value and the radius of the annular fixture are as follows:
[0074] S201. Set the monitoring cycle T. The uniform speed inspection mode means that the infrared sensor moves at a fixed angle S. θ For interval sampling, a uniformly rotating VP performs a patrol inspection, collecting the distance at each location and treating each location as a sampling point; an infrared sensor collects the real-time distance d at each sampling point. current The instantaneous runout deviation of the current sampling point corresponding to the stirring shaft surface relative to the reference center is calculated using the inner radius of the annular fixture and the radius of the reference circle. The formula is: E=|(R ring -d current E represents the instantaneous fluctuation deviation. By setting the monitoring period T and the uniform speed inspection mode, the infrared sensor collects real-time distance values at fixed angular intervals and with uniform rotation, enabling continuous monitoring of the agitator's operating status and timely capture of instantaneous position changes of the agitator shaft. Parameters such as the monitoring period T and the acquisition angle interval can be flexibly adjusted to adapt to different monitoring scenarios based on the operating characteristics of the reactor agitator and actual monitoring needs, thus improving the versatility of the solution.
[0075] S300: Preset jump deviation threshold, use the jump deviation threshold to judge the instantaneous jump deviation to obtain potential abnormal areas, calculate the area severity index of potential abnormal areas, and use the area severity index to set the monitoring priority of all potential abnormal areas;
[0076] The specific steps for setting the monitoring priority of all potentially abnormal areas using the regional severity index are as follows:
[0077] S301, Preset fluctuation deviation threshold E threshold Within a sampling period, when the instantaneous fluctuation deviation E of consecutive sampling points > fluctuation deviation threshold E thresholdWhen a continuous angular region is marked as a "potential abnormal region", the number of continuous sampling points is greater than 1; a preset jump deviation threshold is used to mark potential abnormal regions by judging whether the instantaneous jump deviation of continuous sampling points exceeds the standard, which avoids misjudgment caused by the error of a single sampling point and improves the accuracy of anomaly identification.
[0078] S302. Calculate the regional severity index for each potential abnormal region identified by the marking. The formula is as follows:
[0079] ;
[0080] In the formula, S represents the regional severity index, and E' represents the average instantaneous jump deviation of all sampling points within the potentially abnormal region. The instantaneous jump deviation standard deviation represents all sampling points, L represents the radians corresponding to the angle of the potential anomaly area, and w1, w2, and w2 represent the weights of the average instantaneous jump deviation, the instantaneous jump deviation standard deviation, and the radians, respectively, set by the staff, with w1+w2+w2=1; a regional severity index S is introduced to comprehensively consider the average instantaneous jump deviation, the deviation standard deviation, and the angle radians of the potential anomaly area, fully reflecting the severity of the anomaly area and avoiding the one-sidedness caused by judging based on only a single indicator.
[0081] All potentially abnormal areas are sorted from highest to lowest severity index, and monitoring priorities are assigned based on this ranking: the area ranked first has the highest monitoring priority, and the priority decreases sequentially from the first ranked area. Prioritizing potentially abnormal areas based on their severity index ensures that monitoring resources are concentrated on areas with high severity and risk, improving monitoring efficiency and avoiding resource waste.
[0082] S400: When a high monitoring priority is detected, the uniform speed inspection mode is interrupted, and intensive scanning of key areas is performed.
[0083] The specific steps for performing intensive scanning of key areas are as follows:
[0084] S401, Preset priority threshold P threshold If a potential anomaly is identified as having high priority, and the monitoring priority of the area is greater than or equal to the priority threshold, the uniform speed inspection mode is interrupted, and a dense scan of the key area is performed. Specifically, the dense scan of the key area involves setting the dense angle α and the key scan period T. important Constraint α θ T important >T; When the monitoring priority of a potential abnormal area is less than the priority threshold, the uniform speed inspection mode continues; when the monitoring priority of a potential abnormal area reaches the threshold, the uniform speed inspection mode is interrupted, and intensive scanning of key areas is started. By reducing the acquisition angle interval α and shortening the key scanning cycle, refined monitoring of abnormal areas is achieved, and more comprehensive abnormal data is obtained.
[0085] S402. If a potential anomaly area spreads or moves during a continuous monitoring cycle, the infrared sensor is used to predict the spread and movement area of the potential anomaly area in the next monitoring cycle based on the direction of spread and movement within the continuous monitoring cycle, and to conduct advance tracking and monitoring. Predicting and tracking the spread and movement area in the next monitoring cycle based on the spread and movement of potential anomalies allows for early detection of anomaly expansion trends, buying time for subsequent fault handling. Intensive scanning is performed only on high-priority anomaly areas, while low-priority areas continue to be inspected at a constant speed. This ensures monitoring accuracy in key areas while avoiding the decrease in monitoring efficiency and increased resource consumption caused by intensive scanning of the entire area.
[0086] S500: Within a key scanning cycle, integrate all sampling points collected by the infrared sensor, use interpolation to reconstruct the approximate axis trajectory of the stirring shaft relative to the reference center coordinates, and calculate the comprehensive runout index within the key scanning cycle.
[0087] The specific steps for calculating the comprehensive fluctuation index within the monitoring period are as follows:
[0088] S501. Store all sampling points of the potential abnormal area in the key area intensive scanning within a key scanning cycle, and use the interpolation algorithm to reconstruct the approximate axis trajectory of the stirring shaft relative to the reference center within the key scanning cycle based on all discrete sampling points; use the interpolation algorithm to reconstruct the approximate axis trajectory based on discrete sampling points, and transform the abstract jumping data into an intuitive trajectory graph, so that the staff can quickly understand the offset pattern and abnormal characteristics of the stirring shaft.
[0089] S502. Calculate the comprehensive runout index within the key scanning period. The comprehensive runout index includes the maximum runout, the average runout, the main runout direction, and the trajectory ellipticity. The maximum runout represents the maximum value of the instantaneous runout deviation of all sampling points within the key scanning period. The average runout represents the average value of the instantaneous runout deviation of all sampling points within the key scanning period.
[0090] The main direction of the jump is as follows: within the key scanning period, the absolute value of the instantaneous jump deviation of all sampling points is removed as the offset vector. The offset vector includes the offset amount and the direction, where the direction represents the direction of the unbalance center during the jump from the reference center. The sum of all offset vectors is calculated, and the direction of the sum of the offset vectors is taken as the main direction of the jump.
[0091] The ellipticity of the trajectory is obtained by calculating the ratio of the major and minor axes of the approximate axis-centered trajectory.
[0092] By calculating comprehensive runout indicators such as maximum runout, average runout, main runout direction, and trajectory ellipticity, the runout of the stirring shaft is reflected from multiple dimensions, which can better reveal the operating status of the stirrer compared to a single indicator.
[0093] S600: Preset dynamic threshold library; use dynamic threshold library to judge comprehensive fluctuation index and perform graded early warning; set fault judgment rules based on historical agitator faults; analyze and judge real-time fault type based on approximate shaft center trajectory.
[0094] The specific steps for comparing the approximate axis center trajectory with faults in the fault type database to determine the real-time fault type are as follows:
[0095] S601. The preset dynamic threshold library includes warning thresholds, alarm thresholds, and shutdown thresholds. The dynamic threshold library is used to determine the comprehensive fluctuation index and then issue graded warnings. Specifically:
[0096] When the average fluctuation exceeds the warning threshold, a Level 1 warning is triggered, and a prompt is displayed on the operation interface, while the device continues to operate.
[0097] When the maximum fluctuation exceeds the alarm threshold, a level two warning is triggered, and an audible and visual alarm is activated.
[0098] When the maximum fluctuation exceeds the shutdown threshold, a Level 3 emergency shutdown command is triggered. The system automatically interlocks, stops the mixer, and locks the equipment to prevent accidental restart. Simultaneously, an alarm is sent to staff. A preset dynamic threshold library triggers different levels of warnings based on comprehensive fluctuation indicators. This avoids resource waste caused by excessive warnings for minor anomalies and safety accidents caused by insufficient warnings for serious anomalies. In the Level 3 warning mechanism, the Level 3 emergency shutdown command automatically interlocks and locks the equipment when a serious malfunction occurs, effectively preventing the fault from escalating and avoiding equipment damage and personnel safety risks. Simultaneously, alarm information is sent to staff to ensure timely handling of the fault.
[0099] S602. Preset ellipse threshold and set fault judgment rules based on historical stirrer faults, specifically:
[0100] When the ellipticity of the approximate axis trajectory is less than the elliptic threshold and the main direction of the jump is less than ±10 degrees in the continuous key scanning cycle, the fault is judged as "static quality imbalance".
[0101] When the ellipticity of the approximate axis trajectory is greater than or equal to the elliptic threshold and the main direction of the jump changes periodically within the continuous key scanning cycle, the fault is judged as "axis bending or dynamic misalignment".
[0102] When the approximate axis trajectory is irregular, the fault is judged as "loose bearing or excessive clearance".
[0103] The intelligent management system for detection data applied to reactor agitators includes a baseline calculation module, a real-time fluctuation module, a fault area identification module, an inspection mode adjustment module, a comprehensive analysis module, and a fault early warning module.
[0104] The reference calculation module is used to set up an annular fixture inside the reactor stirrer, deploy an infrared sensor on the annular fixture, use the infrared sensor to collect the initial distance sequence when the stirrer is working, and process the initial distance sequence to obtain the reference distance sequence, reference center coordinates and reference circle radius;
[0105] The real-time jumping module is used to set the monitoring cycle and uniform speed inspection mode when the stirrer is working. The infrared sensor collects the real-time distance value according to the monitoring cycle, and calculates the instantaneous jumping deviation using the real-time distance value and the radius of the annular tool.
[0106] The fault area identification module is used to preset the fluctuation deviation threshold, use the fluctuation deviation threshold to judge the instantaneous fluctuation deviation to obtain potential abnormal areas, calculate the area severity index of the potential abnormal areas, and use the area severity index to set the monitoring priority of all potential abnormal areas.
[0107] The inspection mode adjustment module is used to interrupt the uniform speed inspection mode and perform intensive scanning of key areas when a high monitoring priority is detected.
[0108] The comprehensive analysis module is used to integrate all sampling points collected by the infrared sensor within a monitoring cycle, reconstruct the approximate axis trajectory of the stirrer's stirring shaft relative to the reference center coordinates using interpolation, and calculate the comprehensive runout index within the monitoring cycle.
[0109] The fault early warning module is used to preset a dynamic threshold library, and after judging the comprehensive fluctuation index using the dynamic threshold library, it performs graded early warning. It sets fault judgment rules based on historical agitator faults and analyzes the approximate shaft center trajectory to judge the real-time fault type.
[0110] The comprehensive analysis module includes a center trajectory unit and a comprehensive bounce index unit;
[0111] The axis center trajectory unit is used to reconstruct the approximate axis center trajectory of the stirring shaft relative to the reference center within the key scanning period based on all discrete sampling points using an interpolation algorithm.
[0112] The comprehensive bounce index unit is used to calculate the maximum bounce, average bounce, main bounce direction, and trajectory ellipticity.
[0113] The fault early warning module includes an early warning unit and a fault type unit;
[0114] The early warning unit is used to preset a dynamic threshold library including early warning thresholds, alarm thresholds and shutdown thresholds, and to perform graded early warnings after judging the comprehensive fluctuation index using the dynamic threshold library.
[0115] The fault type unit is used to preset the ellipse threshold and set fault judgment rules based on historical stirrer faults.
[0116] Example:
[0117] For a chemical company's 500L reactor equipped with a paddle agitator (rated speed 60r / min, agitator shaft diameter 80mm), this agitator is used for resin synthesis reaction. During operation, it is necessary to strictly control the runout deviation of the agitator shaft to avoid uneven material mixing, increased equipment wear, or even leakage risks caused by shaft misalignment.
[0118] Hardware configuration: Custom-made annular tooling adapted to the inner wall of the reactor, with an inner radius R. ring =300mm, made of 304 stainless steel, installed in the middle section of the stirring shaft corresponding to the inner wall of the reactor, ensuring that the coaxiality error between the center of the annular tool and the theoretical center of the stirring shaft is ≤0.1mm.
[0119] A laser infrared rangefinder with an accuracy of ±0.01mm and a measurement distance range of 50-500mm is selected and fixed on a ring-shaped fixture. The sensor's measurement direction is pointed towards the center of the stirring shaft, and the response time is ≤1ms.
[0120] Start the stirrer and wait for the speed to stabilize at the rated speed of 60 r / min. Then, control the infrared sensor to rotate uniformly around the annular fixture for one revolution (rotation speed V). P =60° / s), with the sampling angle interval set to θ=0.5° (satisfying the requirement of θ≤1°), a total of n=360 / 0.5=720 initial distances were collected, resulting in the initial distance sequence D0={d1, d2, ..., d 720}, where d1=220.12mm, d2=220.10mm, ..., d 720 =220.11mm.
[0121] The initial distance sequence is processed using a mean filtering algorithm, and the average of the five data points before and after each data point is taken as the baseline distance, resulting in the baseline distance sequence D. base The processed data fluctuation range is ≤ ±0.02mm.
[0122] The reference distance sequence is converted into radial distances from the center of the annular tool to the surface of the stirring shaft, for example, ρ1=300-220.12=79.88mm, ρ2=300-220.10=79.90mm, ..., ρ 720 =300-220.11=79.89mm.
[0123] Based on the angle at each position (the angle at the i-th position is (i-1)×0.5°), calculate the Cartesian coordinates (x, y, y) of the corresponding point on the stirring shaft surface. i y i The formula is x i =ρ i ×cos[(i-1)×0.5°×π / 180], y i =ρ i ×sin[(i-1)×0.5°×π / 180].
[0124] Let the equation of the circle be (x-x0)²+(y-y0)²=R², which can be transformed into the linear equation x²+y²=Ax+By+C. Substituting the 720 Cartesian coordinates into the equation, the least squares solution yields A=0.08mm, B=0.06mm, and C=6382.45mm².
[0125] Substituting A=2x0, B=2y0, and C=R²-x0²-y0², we can calculate the coordinates of the reference center (x0=0.04mm, y0=0.03mm) and the radius of the reference circle R=79.90mm.
[0126] The monitoring cycle is set to T=10s (i.e., a full circumference inspection is completed every 10s). In uniform speed inspection mode, the infrared sensor collects angle intervals S. θ =1°, rotational speed V P =36° / s (ensuring 360° rotation is completed within 10 seconds), real-time distance d of 360 sampling points is collected in each monitoring cycle. current .
[0127] Real-time distance d of a certain sampling point current =220.05mm, and the instantaneous runout deviation is calculated according to the formula E=|(300-220.05)-79.90|=|79.95-79.90|=0.05mm.
[0128] Preset fluctuation deviation threshold E threshold =0.10mm (set according to equipment operating requirements and historical fault data). Within a certain monitoring period, the instantaneous fluctuation deviations of 5 consecutive sampling points (corresponding to an angle range of 5°) are 0.12mm, 0.13mm, 0.11mm, 0.14mm, and 0.12mm, respectively, all greater than E. threshold The 5° angle region is marked as "Potential Anomaly Region A"; the instantaneous fluctuation deviations of three consecutive sampling points (corresponding to an angle range of 3°) are 0.11mm, 0.105mm, and 0.115mm, respectively, and are marked as "Potential Anomaly Region B".
[0129] With preset weights w1=0.5, w2=0.3, w3=0.2 (satisfying w1+w2+w3=1), calculate the regional severity index S for the two potential anomaly regions respectively:
[0130] Potential anomaly region A: S=0.0826; Potential anomaly region B: S=0.0667;
[0131] Sort by regional severity index from largest to smallest: Potentially abnormal region A (0.0826) > Potentially abnormal region B (0.0667), therefore the monitoring priority is: A is level one, and B is level two.
[0132] Preset priority threshold P threshold =0.07. After determining the potential abnormal area A, the uniform speed inspection mode was interrupted, the dense angle α was set to 0.2°, and the key scanning cycle T was adjusted. important =5s, the infrared sensor rotates at 0.2° intervals and a speed V within the potential anomaly region A (5° angle range). P A dense scan is performed at a rate of 1° / s, with 25 sampling points collected in each key scan cycle.
[0133] Within three consecutive monitoring cycles (30s), the potential abnormal area A was found to spread 2° clockwise, expanding the angle range to 7°. Based on the direction of spread, it was predicted that it would continue to spread 1° in the next monitoring cycle (10s). The dense scanning range was expanded to 8° in advance to achieve tracking and monitoring of the abnormal area.
[0134] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.
Claims
1. A method for intelligent management of detection data applied to agitators in a reactor, characterized in that: The method includes the following steps: S100. Set up an annular fixture inside the stirrer of the reactor, deploy an infrared sensor on the annular fixture, use the infrared sensor to collect the initial distance sequence when the stirrer is working, process the initial distance sequence to obtain the reference distance sequence, reference center coordinates and reference circle radius. S200: When the agitator is working, set the monitoring cycle and uniform speed inspection mode. The infrared sensor collects the real-time distance value according to the monitoring cycle, and calculates the instantaneous jump deviation using the real-time distance value and the radius of the ring tool. S300: Preset jump deviation threshold, use the jump deviation threshold to judge the instantaneous jump deviation to obtain potential abnormal areas, calculate the area severity index of potential abnormal areas, and use the area severity index to set the monitoring priority of all potential abnormal areas; S400: When a high monitoring priority is detected, the uniform speed inspection mode is interrupted, and intensive scanning of key areas is performed. S500: Reconstruct the approximate axis trajectory of the stirring shaft relative to the reference center coordinates using interpolation, and calculate the comprehensive runout index within the key scanning cycle; S600: Preset dynamic threshold library; use dynamic threshold library to judge comprehensive fluctuation index and perform graded early warning; set fault judgment rules based on historical agitator faults; analyze and judge real-time fault type based on approximate shaft center trajectory.
2. The intelligent management method for detection data applied to a reactor agitator according to claim 1, characterized in that: The specific steps of performing intensive scanning of key areas in S400 are as follows: Preset priority threshold P threshold When the monitoring priority of a potential anomaly area is greater than or equal to the priority threshold, it is judged as a high priority, the uniform speed inspection mode is interrupted, and a dense scan of the key area is performed. The dense scan of the key area specifically involves setting the dense angle α and the key scan period T. important Constraint α θ T important >T; When the monitoring priority of a potentially abnormal area is less than the priority threshold, continue to execute the uniform speed inspection mode; If a potential abnormal area spreads or moves during a continuous monitoring period, infrared sensors are used to predict the area of potential abnormal area spread and moves in the next monitoring period based on the direction of spread and movement of the potential abnormal area within the continuous monitoring period, and to conduct advance tracking and monitoring.
3. The intelligent management method for detection data applied to a reactor agitator according to claim 1, characterized in that: The S500 comprehensive runout index includes maximum runout, average runout, main runout direction, and trajectory ellipticity; the maximum runout represents the maximum value of the instantaneous runout deviation of all sampling points within the key scanning period; the average runout represents the average value of the instantaneous runout deviation of all sampling points within the key scanning period; The main direction of the jump is as follows: within the key scanning period, the absolute value of the instantaneous jump deviation of all sampling points is removed as the offset vector. The offset vector includes the offset amount and the direction, where the direction represents the direction of the unbalance center during the jump from the reference center. The sum of all offset vectors is calculated, and the direction of the sum of the offset vectors is taken as the main direction of the jump. The ellipticity of the trajectory is obtained by calculating the ratio of the major and minor axes of the approximate axis-centered trajectory.
4. The intelligent management method for detection data applied to a reactor agitator according to claim 1, characterized in that: The fault judgment rules in S600 are as follows: A preset ellipse threshold is used, and fault judgment rules are set based on historical stirrer failures. Specifically: When the ellipticity of the approximate axis trajectory is less than the elliptic threshold and the main direction of the jump is less than ±10 degrees in the continuous key scanning cycle, the fault is judged as "static quality imbalance". When the ellipticity of the approximate axis trajectory is greater than or equal to the elliptic threshold and the main direction of the jump changes periodically within the continuous key scanning cycle, the fault is judged as "axis bending or dynamic misalignment". When the approximate axis trajectory is irregular, the fault is judged as "loose bearing or excessive clearance".
5. The intelligent management method for detection data applied to a reactor agitator according to claim 1, characterized in that: In step S100, the initial distance sequence is processed to obtain the reference distance sequence, reference center coordinates, and reference circle radius as follows: After the reactor agitator is started and reaches its rated speed, an infrared sensor is used to rotate uniformly around the annular fixture for one revolution, collecting initial distance sequences D0={d1, d2, ..., dn} at fixed angle intervals of θ. n }, d1, d2, ..., d n The initial distances represent the distances at the 1st, 2nd, ..., nth positions acquired by the infrared sensor, where each initial distance represents the distance between the infrared sensor and the stirring shaft of the stirrer. A baseline distance sequence D is obtained by processing the initial distance sequence using a mean filtering algorithm. base ; θ≤1°; Using the inner radius R of the ring tool ring The distance value measured by the infrared sensor is converted into polar coordinates of a point on the surface of the stirring shaft, using the formula: ρ i =R ring -d i , where ρ i d represents the radial distance from the center of the annular tool at the i-th position to the surface of the stirring shaft. i This represents the baseline distance at the i-th position in the baseline distance sequence; By collecting the angle at each location and combining it with the corresponding radial distance, the Cartesian coordinates (x, y, z) of the point on the surface of the stirring shaft at each location are calculated. i y i ); Let the equation of the circle to be fitted be given. Substitute the Cartesian coordinates of all stirring shaft surface points transformed from the reference distance sequence into (x, y) of the circle equation. Use the least squares solution to calculate the reference center coordinates (x0, y0) and the reference circle radius R.
6. The intelligent management method for detection data applied to a reactor agitator according to claim 1, characterized in that: The uniform speed inspection mode in S400 means that the infrared sensor moves at a fixed angle S. θ For interval sampling, the VP rotates at a constant speed to perform inspection, collecting the distance at each position and treating each position as a sampling point.
7. The intelligent management method for detection data applied to a reactor agitator according to claim 1, characterized in that: The S400 mechanism uses a regional severity index to set the monitoring priority for all potentially abnormal areas as follows: Preset fluctuation deviation threshold E threshold Within a sampling period, when the instantaneous fluctuation deviation E of consecutive sampling points > fluctuation deviation threshold E threshold When a continuous angular region is marked as a "potential anomaly region", the number of continuous sampling points is greater than 1. For each potential abnormal region identified by the label, a regional severity index is calculated. All potential abnormal regions are then sorted from largest to smallest based on their regional severity indices. Monitoring priorities are set according to the sorting, specifically: the region ranked first has the highest monitoring priority, and the monitoring priorities decrease sequentially according to the sorting.
8. An intelligent management system for detection data applied to agitators in a reactor, characterized in that: The intelligent management system for detection data includes a benchmark calculation module, a real-time fluctuation module, a fault area identification module, an inspection mode adjustment module, a comprehensive analysis module, and a fault early warning module. The reference calculation module is used to set up an annular fixture inside the reactor stirrer, deploy an infrared sensor on the annular fixture, use the infrared sensor to collect the initial distance sequence when the stirrer is working, and process the initial distance sequence to obtain the reference distance sequence, reference center coordinates and reference circle radius; The real-time jumping module is used to set the monitoring cycle and uniform speed inspection mode when the stirrer is working. The infrared sensor collects the real-time distance value according to the monitoring cycle, and calculates the instantaneous jumping deviation using the real-time distance value and the radius of the annular tool. The fault area identification module is used to preset the fluctuation deviation threshold, use the fluctuation deviation threshold to judge the instantaneous fluctuation deviation to obtain potential abnormal areas, calculate the area severity index of the potential abnormal areas, and use the area severity index to set the monitoring priority of all potential abnormal areas. The inspection mode adjustment module is used to interrupt the uniform speed inspection mode and perform intensive scanning of key areas when a high monitoring priority is detected. The comprehensive analysis module is used to integrate all sampling points collected by the infrared sensor within a monitoring cycle, reconstruct the approximate axis trajectory of the stirrer's stirring shaft relative to the reference center coordinates using interpolation, and calculate the comprehensive runout index within the monitoring cycle. The fault early warning module is used to preset a dynamic threshold library, and after judging the comprehensive fluctuation index using the dynamic threshold library, it performs graded early warning. It sets fault judgment rules based on historical agitator faults and analyzes the approximate shaft center trajectory to judge the real-time fault type.
9. The intelligent management system for detection data applied to a reactor agitator according to claim 8, characterized in that: The comprehensive analysis module includes an axis trajectory unit and a comprehensive jump index unit; The axis center trajectory unit is used to reconstruct the approximate axis center trajectory of the stirring shaft relative to the reference center within the key scanning period based on all discrete sampling points using an interpolation algorithm. The comprehensive bounce index unit is used to calculate the maximum bounce, average bounce, main bounce direction, and trajectory ellipticity.
10. The intelligent management system for detection data applied to a reactor agitator according to claim 8, characterized in that: The fault early warning module includes an early warning unit and a fault type unit; The early warning unit is used to preset a dynamic threshold library including early warning thresholds, alarm thresholds and shutdown thresholds, and to perform graded early warnings after judging the comprehensive fluctuation index using the dynamic threshold library. The fault type unit is used to preset the ellipse threshold and set fault judgment rules based on historical stirrer faults.