Early warning device and method for liquid leakage of cast aluminum bar in deep well

Abstract

According to the deep well casting aluminum bar liquid leakage early warning device and method, the multiple temperature sensors are evenly arranged on the periphery of an aluminum bar stretching outlet, multi-point real-time collection of circumferential temperature information of a formed aluminum bar can be effectively achieved, limitation of single-point detection is avoided, and the accuracy of liquid leakage early warning is improved through hierarchical noise reduction processing of upper computer software. A bilinear interpolation algorithm is adopted to generate a two-dimensional matrix of continuous temperature distribution in the circumferential direction of the aluminum bar, the upper computer renders the two-dimensional matrix into a dynamic thermodynamic diagram and displays the dynamic thermodynamic diagram on an interface, and when the temperature exceeds a safety range, an early warning is given out immediately. A temperature data real-time analysis and threshold temperature judgment system is constructed, temperature field gradient mutation characteristics caused by liquid leakage can be accurately captured, false alarms caused by environmental noise such as mechanical vibration and steam interference are effectively restrained, meanwhile, a single-point monitoring blind area is eliminated through a multi-point collaborative temperature measurement mechanism, and the risk of missed alarms is avoided.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of metal processing safety monitoring, in particular to a deep-well cast aluminum rod liquid leakage early warning device and method. BACKGROUND

[0002] In the casting process, molten aluminum enters the mold through the pouring system, and after solidification, the desired shape of the casting is formed, but in the deep-well aluminum processing in the factory, due to factors such as mold wear and uneven solidification shrinkage during aluminum rod stretching and forming, small liquid leaks often occur without being easily detected. Such leaks not only cause structural, quality and functional problems in the casting, but more dangerously, the high-temperature aluminum liquid will cause a violent gasification reaction in the instant of contacting the cooling water system, producing a local high-temperature and high-pressure steam shock wave, thereby causing serious safety accidents.

[0003] Current aluminum liquid leakage detection technology mainly monitors liquid level fluctuations through an image recognition system or uses a single-point thermocouple to collect outlet end surface temperature, but due to factors such as steam interference and mechanical device vibration in the casting environment, the visual system has a high misjudgment rate; the traditional single-point temperature measurement method is limited by the hysteresis of heat conduction, making it difficult to capture small liquid leaks that occur in the early stage of leakage, and the existing system lacks dynamic analysis capability for temperature gradient mutations in the leakage area, resulting in the leakage amount exceeding the critical third safety threshold when the alarm is triggered, affecting the efficiency of the entire production line, increasing the cost of quality inspection and rework, and causing serious safety accidents.

[0004] In view of this, the present application discloses a deep-well cast aluminum rod liquid leakage early warning device and method. SUMMARY

[0005] To solve the above problems, the technical scheme of the present application is as follows: The present application discloses a deep-well cast aluminum rod liquid leakage early warning method, which comprises the following steps: Temperature measurement is performed on multiple points around the aluminum rod, temperature data analysis is performed, and hierarchical denoising processing is performed; A bilinear interpolation algorithm is performed on the denoised discrete temperature data to generate a two-dimensional matrix of the circumferential continuous temperature distribution of the aluminum rod (4); The two-dimensional temperature matrix is rendered into a dynamic heat map, an incremental color mapping is adopted, the gradient mutation area is automatically marked and highlighted with a flashing warning, wherein: the blue system represents normal temperature, and the red system represents early warning threshold temperature, and the threshold temperature includes a first safety threshold, a second safety threshold and a third safety threshold; A hierarchical early warning mechanism is adopted: When the single-point temperature exceeds the first safety threshold, a first-level early warning is triggered; When the temperature difference between two adjacent points around the aluminum rod exceeds the second safety threshold and lasts for 2 seconds, a second-level early warning is triggered; When the temperature exceeds the preset third safety threshold for 5 seconds, a multi-stage linkage alarm is triggered.

[0006] Preferably, the hierarchical noise reduction processing includes the following steps: High-frequency mechanical noise is suppressed by sliding average filtering, and random fluctuations are weakened by using temperature data mean value; An adaptive Kalman filter algorithm is introduced to dynamically estimate noise covariance and dynamically adjust filter gain, and a multi-scale decomposition is used to denoise the temperature signal; Combined with Db4 wavelet basis three-layer decomposition, the temperature signal is decomposed into approximate components and detail components, and each detail component is processed by an adaptive threshold function, and then the processed components are reconstructed to eliminate non-stationary noise caused by steam interference.

[0007] Preferably, the bilinear interpolation algorithm includes the following steps: An polar coordinate system is established with the center of the aluminum rod as the origin, and the position coordinates are calibrated; The circumference of the aluminum rod is divided into N equal angles, and the radius is divided into M equal radii to form a two-dimensional grid; The temperature of the grid point is calculated according to the temperature data using the bilinear interpolation formula; The temperature value of the fixed point on the surface of the aluminum rod is calculated, the discrete temperature sampling point data is converted into continuous temperature distribution, and a two-dimensional temperature matrix is generated; A Gaussian filter kernel is used on the two-dimensional temperature matrix, and the data points in the temperature matrix are weighted and averaged to realize smoothing processing of the two-dimensional temperature matrix.

[0008] Preferably, the dynamic thermal map further includes: The real-time temperature curve and the historical normal temperature curve are superimposed on the same screen, A time axis slider is set to query historical data, and a third safety threshold prompt box is marked; The temperature gradient change rate and the threshold deviation are displayed when the alarm is triggered.

[0009] Preferably, the present scheme further provides a deep well cast aluminum rod liquid leakage early warning device, comprising: A rectangular array temperature measurement mold includes a plurality of rectangular temperature measurement modules arranged in an array for accommodating stretched aluminum rods; A plurality of temperature sensors are arranged at 90° angles around each rectangular temperature measurement module, and a plurality of temperature sensors are configured with independent electromagnetic shielding pipes; A temperature transmitter receives the electrical signal output by the temperature sensor and converts it into a standard electrical signal that is linearly related to the temperature value; The host computer receives a standard electrical signal corresponding to a temperature value in a linear relationship transmitted by the temperature transmitter, and analyzes the standard electrical signal, and sequentially performs hierarchical noise reduction, bilinear interpolation temperature field reconstruction, and renders the reconstructed two-dimensional temperature matrix into a dynamic heat map and visually displays it, thereby prewarning the aluminum bar liquid leakage.

[0010] Preferably, the diameter of the central through hole of the rectangular temperature measuring module is 10mm-15mm larger than the diameter of the aluminum bar.

[0011] Preferably, the surface of the rectangular array temperature measuring mold is plasma sprayed with a heat insulation coating layer with a thickness of 0.8mm-1.2mm.

[0012] Preferably, the rectangular temperature measuring module is internally provided with a cavity, and the temperature sensor is centrally wired through the cavity in the rectangular temperature measuring module.

[0013] The technical scheme provided by the present application has the following beneficial effects: The present application can effectively realize the multi-point real-time acquisition of the circumferential temperature information of the formed aluminum bar by uniformly arranging a plurality of temperature sensors around the aluminum bar drawing outlet, avoids the limitation of single-point detection, adjusts the filter gain dynamically through the hierarchical noise reduction processing of the host computer, performs multi-scale decomposition and denoising on the temperature signal, generates a two-dimensional matrix of the circumferential continuous temperature distribution of the aluminum bar by using the bilinear interpolation algorithm, renders the two-dimensional matrix into a dynamic heat map by the host computer, and displays it on the interface, immediately issues a prewarning when the temperature exceeds the safety range, and builds a temperature data real-time analysis and threshold temperature discrimination system, which can accurately capture the temperature field gradient mutation characteristics caused by liquid leakage, effectively suppresses the false alarm caused by environmental noise such as mechanical vibration and steam interference, eliminates the single-point monitoring blind area through the multi-point cooperative temperature measurement mechanism, and avoids the risk of missing the alarm. BRIEF DESCRIPTION OF DRAWINGS

[0014] The accompanying drawings, which are included to provide a further understanding of the present application, constitute a part of the present application, and the illustrative embodiments of the present application and their descriptions serve to explain the present application and do not constitute an improper limitation on the present application.

[0015] Wherein: Figure 1 is a structural schematic diagram of the rectangular array temperature measuring mold of the present application; Figure 2 is a top view of the rectangular array temperature measuring mold of the present application; Figure 3 is an enlarged structural schematic diagram of the rectangular temperature measuring module of the present application; Figure 4 is a hardware system flow chart of the liquid leakage prewarning device of the present application; Figure 5 is a data processing flow chart of the present application.

[0016] Label explanation: 1, rectangular array temperature measurement mold; 2, rectangular temperature measurement module; 3, temperature sensor; 4, aluminum bar. DETAILED DESCRIPTION

[0017] In order to make the technical problems to be solved by the present application, technical solutions and beneficial effects more clear, specific embodiments of the present application will be further described in detail below. It should be understood that the specific embodiments described herein are only used to explain the present application and not used to limit the present application.

[0018] Reference Figures 1-5 As shown, the embodiment of the present application provides a deep well casting aluminum bar liquid leakage early warning method, specifically, comprising the following steps: The temperature sensor 3 measures the temperature, the temperature transmitter transmits to the upper computer through RS485, the upper computer analyzes the temperature data based on the Modbus protocol, and performs hierarchical denoising processing, the data processing adopts hierarchical denoising strategy, first, the high-frequency mechanical noise is suppressed by sliding average filtering, and the random fluctuation is weakened by using the data mean value in the sliding window, such as L = 15, which corresponds to suppressing 5Hz-10Hz high-frequency noise, when new data enters, the original data in the window is removed and the mean value is calculated. Based on the dynamic characteristics of the casting environment, the adaptive Kalman filter algorithm is introduced, the temperature signal is regarded as the state quantity to establish the state space model, the system is described by the state transition equation and the observation equation, and the process noise covariance and the observation noise covariance are dynamically adjusted to adjust the Kalman gain After the initialization parameters are predicted, the noise covariance is adaptively updated, and the step is updated, the time-varying noise and system uncertainty are processed, that is, the Kalman filter estimates the noise covariance matrix in real time, dynamically adjusts the filter gain, and performs multi-scale decomposition and denoising on the temperature signal, realizes the robustness improvement of non-stationary noise, and combines with the Db4 wavelet base three-layer decomposition to remove the non-stationary noise caused by steam interference. The multi-resolution analysis of the Db4 wavelet decomposes the temperature signal into an approximate component and a detail component, and three-layer decomposition obtains , each detail component is processed by an adaptive threshold function, and the processed component is reconstructed to retain the temperature mutation characteristics;

[0019] The bilinear interpolation algorithm is performed on the denoised discrete temperature data to generate a two-dimensional matrix of the circumferential continuous temperature distribution of the aluminum bar 4. In the visualization interface, the original data collected is expanded to the circumferential continuous temperature distribution data of the aluminum bar 4 by the interpolation algorithm after analysis. The bilinear interpolation method is used with the sensor position as the reference point, and the specific processing is as follows:

[0020] Raw temperature data of PT100 sensors is read, corresponding to evenly distributed points (such as 0°, 90°, 180°, 270°) at 90° equiangular positions of the circumference of the aluminum bar 4, the sensor position coordinates are calibrated, and a polar coordinate system is established with the center of the aluminum bar 4 as the origin, and the sensor positions are converted to , wherein r is the distance from the mold to the center point of the aluminum bar 4, is the angle, the circumference of the aluminum bar 4 is divided into N equally spaced angles, and the radial direction is divided into M equally spaced radii to form a two-dimensional grid. Using a bilinear interpolation formula, the temperature of each grid point is calculated according to the temperature value of each vertex sensor . By calculating the temperature values of the fixed points on the surface of the aluminum bar 4, the discrete temperature sampling point data is converted into a continuous temperature distribution, and a two-dimensional temperature matrix is generated. A 5x5 Gaussian filter kernel is used for the two-dimensional temperature matrix, which uses a Gaussian function to perform weighted averaging on the data points in the temperature matrix, achieving smoothing processing of the two-dimensional temperature matrix to reduce noise and temperature fluctuations caused by mechanical vibration, sensor errors, etc. The scheme uses a polar coordinate system for grid division, making the interpolation grid more matched to the temperature distribution characteristics of the circumference of the aluminum bar 4. The equidistant division of the circumferential angle ensures the uniform reconstruction of the temperature field in the circumferential direction, avoiding edge distortion caused by using traditional Cartesian coordinates for conversion. Taking the actual deployment position of the sensor as the reference point, the sensor coordinates are calibrated to ensure that the reference points for interpolation calculation strictly correspond to the hardware layout. In the polar coordinate system, the (r, θ) coordinates of each PT100 sensor are labeled to avoid temperature field reconstruction errors caused by coordinate deviations. Advantage: The interpolation algorithm focuses on enhancing the ability to capture temperature gradient changes in response to the local temperature mutation characteristics caused by liquid leakage in deep well casting. When performing Gaussian filtering smoothing processing after generating the temperature matrix, the edge features of the gradient mutation region are preserved, so that the interpolated temperature field can reflect both the overall distribution and the abnormal temperature rise of the leakage point,

[0021] and is intuitively displayed in the form of a heat map. The progressive color mapping scheme is adopted, with blue for normal operating temperature, red for temperature close to or exceeding the warning threshold, and intermediate for temperature gradient transition. The color mapping is as follows: set the temperature range , and normalize any temperature T to the interval [0, 1], where , and according to tThe difference value obtains the corresponding color in the color scale. The temperature matrix is rendered in real time as a dynamic heat map, which is displayed in real time on the host computer interface, and the surface temperature distribution of the aluminum bar 4 is intuitively displayed. The circumferential temperature distribution of the mold is rendered in real time, and the gradient mutation area is automatically marked. When the temperature abnormal area appears, the heat map is automatically highlighted, and the warning is given in the form of flickering. At the same time, the real-time temperature curve and the historical normal temperature curve are superimposed and displayed on the same screen, and the deviation of the current temperature from the normal state is intuitively presented. The time axis slider is set in the interface to query the historical temperature data. The operator can quickly locate the temperature change trend in a specific time period by dragging the slider, and view the temperature fluctuation before and after the liquid leakage warning event occurs. The preset safety temperature threshold is displayed in the form of a prompt box in the temperature curve, and multiple sets of historical data or real-time data at different positions can be compared, which is beneficial to analyze the temperature change difference of different time periods and different positions. When the difference between the real-time temperature data before and after is higher than the safety temperature threshold, the warning mechanism is triggered, and the warning information is displayed in the center of the interface, indicating the current temperature gradient change rate, threshold, deviation degree and operation measures;

[0022] A hierarchical early warning mechanism is adopted: When the single-point temperature changes suddenly, the temperature difference of the adjacent sensor is checked. When the single-point temperature exceeds the first safety threshold, a first-level warning is triggered, indicating that the local temperature rise is abnormal and there is a potential risk of liquid leakage. When the temperature difference of the adjacent sensor exceeds the second safety threshold and lasts for 2 seconds, a second-level warning is triggered. A real-time temperature curve is drawn, and a multi-level cascading alarm is triggered within 5 seconds when the temperature exceeds the preset threshold. This scheme realizes real-time capture of early-stage temperature abnormalities of liquid leakage through multi-point collaborative temperature measurement, anti-interference hardware design, hierarchical noise reduction strategy and first safety threshold algorithm. The spatial sampling advantage and anti-interference design of the hardware device provide accurate raw data for data processing, and the two form a closed loop of "collection-processing-decision", realizing real-time capture of early-stage temperature abnormalities of liquid leakage, solving the hysteresis problem of traditional detection, and significantly improving the accuracy and timeliness of the warning, providing an efficient and reliable technical path for safety monitoring of deep well casting.

[0023] Please refer to Figures 1-5 , which is a deep well casting aluminum bar liquid leakage warning device as the best embodiment of the present application, comprising: A rectangular array temperature measurement mold 1 comprises a plurality of rectangular temperature measurement modules 2 arranged in an array for accommodating a stretched aluminum bar 4. The rectangular array temperature measurement mold 1 is a customized anti-magnetic high-temperature-resistant rectangular mold, and the central through hole has a diameter of 10mm-15mm larger than the specification of the aluminum bar 4, which can ensure that the stretched aluminum bar 4 passes through the center without contact. The mold is made of silicon carbide-nitride boron composite ceramic material, and the surface is coated with 0.8mm-1.2mm zirconia thermal barrier coating by plasma spraying; A plurality of temperature sensors 3 are arranged at 90° intervals around each rectangular temperature measurement module 2, and a plurality of temperature sensors 3 are configured with independent electromagnetic shielding conduits, wherein four groups of PT100 patch type temperature sensors 3 are embedded at 90° intervals around each rectangle in the mold; A temperature transmitter receives the temperature data detected by the temperature sensor 3 through a cable, the cable of the temperature sensor 3 is centrally routed through the center cavity of the mold, and after being led out through an insulation layer, it is connected to a PT100 temperature transmitter to receive raw data, which is transmitted to the upper computer through RS485 and is parsed based on the Modbus protocol, thereby warning of aluminum rod 4 liquid leakage.

[0024] In summary, the present application uniformly arranges a plurality of temperature sensors 3 around the stretching outlet of the aluminum rod 4, which can effectively realize the multi-point real-time acquisition of the circumferential temperature information of the formed aluminum rod 4, avoids the limitations of single-point detection, and dynamically adjusts the filter gain based on the hierarchical noise reduction processing of the Modbus protocol of the upper computer software, performs multi-scale decomposition and denoising on the temperature signal, generates a two-dimensional matrix of the circumferential continuous temperature distribution of the aluminum rod 4 by using the bilinear interpolation algorithm, renders it into a dynamic heat map by the upper computer, and displays it on the interface, immediately issues a warning when the temperature exceeds the safety range, and constructs a temperature data real-time analysis and threshold temperature discrimination system, which can accurately capture the temperature field gradient mutation characteristics caused by liquid leakage, effectively suppress false positives caused by environmental noise such as mechanical vibration and steam interference, eliminate single-point monitoring blind spots through a multi-point cooperative temperature measurement mechanism, and avoid the risk of missing reports.

[0025] The above describes the present application in conjunction with the drawings, and it is obvious that the specific implementation of the present application is not limited to the above manner, as long as various non-essential improvements are made by using the method concept and technical solution of the present application, or the concept and technical solution of the present application is directly applied to other occasions without improvement, all of which are within the protection scope of the present application.

Claims

1. A method for early warning of leakage in deep-well cast aluminum rods, characterized in that, Includes the following steps: Temperature measurements were taken at multiple points around the aluminum rod (4), the temperature data were analyzed, and graded noise reduction was performed. A bilinear interpolation algorithm is performed on the denoised discrete temperature data to generate a two-dimensional matrix of the continuous circumferential temperature distribution of the aluminum rod (4); The two-dimensional temperature matrix is ​​rendered as a dynamic heat map, and a progressive color mapping is used to automatically mark gradient change areas and highlight and flash warnings. Among them, blue represents normal temperature and red represents warning threshold temperature. The threshold temperature includes a first safety threshold, a second safety threshold and a third safety threshold. A tiered early warning mechanism is adopted: When the temperature at a single point exceeds the first safety threshold, a Level 1 warning is triggered. When the temperature difference between two adjacent points in the circumferential direction of the aluminum rod (4) exceeds the second safety threshold and lasts for 2 seconds, a level 2 warning is triggered. When the temperature exceeds the preset third safety threshold for 5 seconds, a multi-level linkage alarm is triggered.

2. The method for early warning of leakage in deep well cast aluminum rods according to claim 1, characterized in that, The graded noise reduction process includes the following steps: High-frequency mechanical noise is suppressed by using moving average filtering, and random fluctuations are weakened by using the mean of temperature data. An adaptive Kalman filter algorithm is introduced to dynamically estimate the noise covariance, dynamically adjust the filter gain, and decompose the temperature signal at multiple scales to remove noise. By combining the three-level decomposition of the Db4 wavelet basis, the temperature signal is decomposed into approximate components and detail components. An adaptive threshold function is used to denoise each detail component, and then the processed components are reconstructed to remove non-stationary noise caused by steam interference.

3. The method for early warning of leakage in deep well cast aluminum rods according to claim 1, characterized in that, The bilinear interpolation algorithm includes the following steps: Establish a polar coordinate system with the center of the aluminum rod (4) as the origin, and calibrate the position coordinates; The aluminum rod (4) is divided into N equal parts of angle in the circumference and M equal parts of radius in the radial direction to form a two-dimensional grid; The temperature at grid points is calculated based on the temperature data using the bilinear interpolation formula. By calculating the temperature values ​​at fixed points on the surface of the aluminum rod (4), the discrete temperature sampling point data is converted into a continuous temperature distribution, generating a two-dimensional temperature matrix; A Gaussian filter kernel is used to apply a weighted average to the data points in the two-dimensional temperature matrix, thereby achieving smoothing of the two-dimensional temperature matrix.

4. The method for early warning of leakage in deep well cast aluminum rods according to claim 1, characterized in that, The dynamic heat map also includes: The real-time temperature curve and the historical normal temperature curve are overlaid on the same screen. Set a timeline slider to query historical data and add a prompt box indicating the third safety threshold. The warning displays the rate of change of the temperature gradient and the threshold deviation.

5. A leakage early warning device for deep-well cast aluminum rods, based on the leakage early warning method for deep-well cast aluminum rods according to any one of claims 1 to 4, characterized in that, include: The rectangular array temperature measuring mold (1) includes several rectangular temperature measuring modules (2) arranged in an array to accommodate the stretched aluminum rod (4). Several temperature sensors (3) are arranged at 90° angles around each rectangular temperature measuring module (2), and each of the temperature sensors (3) is equipped with an independent electromagnetic shielding conduit. The temperature transmitter receives the electrical signal output by the temperature sensor (3) and converts it into a standard electrical signal that is linearly related to the temperature value. The host computer receives and analyzes the standard electrical signal transmitted by the temperature transmitter that is linearly related to the temperature value. It then performs graded noise reduction, bilinear interpolation temperature field reconstruction, and renders the reconstructed two-dimensional temperature matrix into a dynamic thermogram for visualization. It also provides an early warning for leakage of aluminum rod (4).

6. The deep well casting aluminum rod leakage early warning device according to claim 5, characterized in that, The diameter of the central through hole of the rectangular temperature measuring module (2) is 10mm to 15mm larger than the diameter of the aluminum rod (4).

7. The deep well casting aluminum rod leakage early warning device according to claim 5, characterized in that, The surface of the rectangular array temperature measuring mold (1) is plasma-sprayed with a heat insulation coating of 0.8mm to 1.2mm.

8. The deep well casting aluminum rod leakage early warning device according to claim 5, characterized in that, The rectangular temperature measuring module (2) has an internal cavity, and the temperature sensor (3) is wired through the cavity inside the rectangular temperature measuring module (2).

Images