Electrode pressure relief measuring device and intelligent smelting system of electric arc furnace

Abstract

The application relates to an electrode pressure relief measuring device and an intelligent smelting system of an electric arc furnace. The device comprises: at least three non-contact displacement sensors which are distributed around the circumference of the same electrode shell, each non-contact displacement sensor being fixedly installed at a different circumferential position of an electrode holder and being used for synchronously collecting displacement signals of the electrode shell at the corresponding circumferential position relative to the electrode holder; and a data processing unit which receives the displacement signals collected by the non-contact displacement sensors and performs the following processing: based on the displacement signals of the at least three circumferential positions, the three-dimensional space posture change of the electrode in the pressure relief process is solved. The device synchronously collects the displacement signals of each position relative to the electrode holder by circumferentially distributing at least three non-contact displacement sensors around the electrode shell, solves the tilting direction and tilting angle of the electrode axis in the pressure relief process, provides a quantitative basis for posture monitoring and abnormal diagnosis, and overcomes the limitation that the traditional single-point measurement cannot perceive the space posture of the electrode.

Description

Technical Field

[0001] This application relates to the field of intelligent equipment technology, and in particular to an electrode pressure measurement device and an intelligent smelting system for an electric arc furnace. Background Technology

[0002] With the continuous development of submerged arc furnace smelting technology, electrode pressing and releasing control is increasingly widely used in the production of ferroalloys, calcium carbide, and other products. The introduction of non-contact displacement sensing and data acquisition technology has provided a technical foundation for real-time monitoring of the electrode pressing and releasing process, making automated pressing and releasing control possible. However, in actual production, electrodes face complex conditions such as high temperature, strong electromagnetic interference, and mechanical vibration during the pressing and releasing process. Existing measurement methods typically use a single or a small number of displacement sensors for single-point measurement, which can only obtain displacement information of a certain local part of the electrode. They lack the ability to coordinate and integrate multiple sensors around the circumference of the electrode and the ability to sense the tilt direction and tilt angle of the electrode axis in three-dimensional space. When the electrode is tilted or locally stuck, single-point measurement data often masks the abnormal posture, making it difficult for the system to accurately determine the true motion state of the electrode.

[0003] In related technologies, the judgment rules for pressure release events are relatively simple, usually using whether the displacement signal of a single channel exceeds a set threshold as the trigger condition. There is no joint judgment mechanism that includes monotonicity detection of all circumferential signals and verification of the total displacement range. Vibration interference, signal noise, or minor local fluctuations are easily misjudged as pressure release actions, leading to inaccurate pressure release recordings and affecting the correct assessment of electrode consumption rates. Simultaneously, the anomaly diagnosis methods are rudimentary. Existing systems cannot distinguish between different anomaly types such as overall slippage and tilting jamming based on the step back characteristics or over-limit characteristics of displacement signals combined with three-dimensional attitude information. Operators lack clear guidance on fault attribution when faced with alarms. Pressure release cycles are typically executed at fixed time intervals, failing to establish a correlation model between historical pressure release sequences and electrical parameters such as electrode current, power, and voltage to dynamically predict consumption rates. This prevents the adjustment of the pressure release rhythm according to real-time operating conditions, resulting in either excessive pressure release leading to electrode waste or insufficient pressure release causing electrode overheating or furnace condition fluctuations. This severely restricts the automation level and operational economy of the submerged arc furnace smelting process. Summary of the Invention

[0004] Therefore, it is necessary to provide an electrode pressure measurement device and an intelligent smelting system for an electric arc furnace to address the above-mentioned technical problems.

[0005] In a first aspect, this application provides an electrode pressure discharge measuring device, comprising: At least three non-contact displacement sensors are circumferentially distributed around the same electrode shell. Each of the non-contact displacement sensors is fixedly installed at different circumferential positions of the electrode holder to synchronously acquire the displacement signal of the electrode shell relative to the electrode holder at the corresponding circumferential position. The data processing unit receives the displacement signals collected by each of the non-contact displacement sensors and performs the following processing: The three-dimensional spatial attitude change of the electrode during the pressing and releasing process is calculated based on the displacement signals of the at least three circumferential positions. The three-dimensional spatial attitude change includes at least the tilt direction and tilt angle of the electrode axis. Extract the variation characteristics of displacement signals at each circumferential position over time. If the displacement signals at all circumferential positions are detected to be monotonically increasing within the preset compression time window, and the total displacement at each circumferential position falls within the preset compression range, it is determined to be a real compression event and the compression amount is recorded. When a step back is detected in the displacement signal of at least one circumferential position outside the preset compression time window, or when the total displacement of the displacement signal of at least one circumferential position within the preset compression time window exceeds the upper limit of the preset compression range, the abnormal event that has occurred is determined to be an overall slippage event or a tilting and jamming event based on the three-dimensional spatial attitude change.

[0006] In one embodiment, calculating the three-dimensional spatial attitude change of the electrode during the pressing and releasing process based on the displacement signals at the at least three circumferential positions includes: The mounting angle of each non-contact displacement sensor in the circumferential direction of the electrode shell and the corresponding real-time displacement value are obtained. Using the electrode holder as a reference, the spatial discrete point coordinates of the bottom end face of the electrode shell at each sampling time are constructed based on the installation angle and real-time displacement value of each sensor. Plane fitting is performed on the coordinates of each of the spatial discrete points to obtain the spatial plane equation of the bottom end face of the electrode shell, and the normal vector of the spatial plane is calculated. The normal vector is compared with the gravity direction vector, and the angle at which the normal vector deviates from the gravity direction is determined as the tilt angle of the electrode axis. The projection direction of the normal vector onto the horizontal plane is taken as the tilt direction.

[0007] In one embodiment, extracting the time-varying characteristics of the circumferential displacement signals includes: Real-time sliding median filtering or mean filtering is performed on the displacement signals of each circumferential position to remove high-frequency noise interference; Perform a first-order difference operation on the filtered displacement signal to obtain the instantaneous displacement change rate at each sampling time. According to the preset time interval or number of sampling points, the filtered displacement signals of each circumferential position are accumulated and summed to obtain the total displacement of each circumferential position within any specified time period. The change characteristics include a real-time displacement value sequence, an instantaneous displacement change rate sequence, and the total displacement within a specified time period.

[0008] In one embodiment, when the displacement signals at all circumferential positions are detected to be monotonically increasing within a preset release time window, and the total displacement at each circumferential position falls within a preset release range, a true release event is determined and the release amount is recorded, including: The moment when the instantaneous displacement change rate of at least one circumferential position is first detected to exceed a preset start threshold is taken as the start time of the compression and release time window, and a preset fixed duration is taken as the length of the compression and release time window. Within the compression and release time window, it is continuously determined whether the difference between the displacement signals of each circumferential position at adjacent sampling times is greater than or equal to zero. If a negative difference value appears in any channel, the determination is terminated. At the end of the pressing and releasing time window, the displacement difference between the end and start times of each circumferential position is calculated and used as the total displacement of that circumferential position in that cycle. The total displacement at each circumferential position in this cycle is compared with the lower and upper limits of the preset compression range, respectively. If the displacement signals of all circumferential positions within the window are monotonically increasing, and the total displacement of each circumferential position in this instance falls within the preset pressure release range, then it is determined to be a real pressure release event, and the average or weighted average of the total displacement of all circumferential positions in this instance is recorded as the pressure release amount for this instance.

[0009] In one embodiment, detecting a step back in the displacement signal at the at least one circumferential position outside the preset pressing and releasing time window includes: Real-time monitoring of displacement signals in each circumferential position; when it is detected that the current moment is not within any triggered release time window, calculate the displacement difference between the current moment and the previous moment for each circumferential position. If the current displacement difference at a certain circumferential position is negative and its absolute value exceeds the preset step back threshold, and the displacement signal continues to decrease over multiple consecutive sampling times and the total cumulative decrease exceeds the preset cumulative back threshold, then it is determined that a step back has occurred at the circumferential position, and the event and the circumferential position where it occurred are recorded.

[0010] In one embodiment, detecting that the total displacement of the displacement signal at the at least one circumferential position exceeds the upper limit of the preset release time window includes: Within each compression and release time window, the displacement of each circumferential position is accumulated in real time, and the real-time accumulated displacement is compared with the upper limit of the preset compression and release range; When the real-time cumulative displacement at any of the aforementioned circumferential positions exceeds the upper limit of the preset pressure release range, an over-range flag is immediately triggered, and the time corresponding to the triggering of the over-range flag and the information of the corresponding circumferential position are recorded. Monitoring continues until the end of the window.

[0011] In one embodiment, determining whether the currently occurring abnormal event is an overall slippage event or a tilting and jamming event based on the three-dimensional spatial posture change includes: When a step back event or a total displacement exceeding the limit event is detected, the tilt angle and tilt direction of the electrode axis at the current moment are obtained; The tilt angle is compared with a preset tilt threshold. If the tilt angle is less than the tilt threshold, and the displacement signals at each circumferential position change in the same direction and with similar amplitudes, then the current abnormal event is determined to be an overall slippage event. If the tilt angle is greater than or equal to the tilt threshold, and there are at least two circumferential positions where the displacement signals change in opposite directions or the displacement difference exceeds a preset inconsistency threshold, then the current abnormal event is determined to be a tilt jamming event.

[0012] Secondly, this application also provides an intelligent smelting system for an electric arc furnace, comprising: The electrode pressure discharging measuring device described above; The smelting electrical parameter acquisition unit is used to acquire electrode current, furnace power and electrode-to-ground voltage in real time. The smelting data processing unit receives the discharge amount data, three-dimensional spatial attitude change and abnormal event judgment results output by the electrode discharge measurement device, and receives the electrical parameter data output by the smelting electrical parameter acquisition unit. The smelting data processing unit establishes a correlation model between electrode consumption rate and electrical parameters based on the pressure release sequence and electrical parameter sequence within a preset historical time window. Using the correlation model, it predicts the expected consumption rate of the electrode in real time based on the current electrical parameters and generates pressure release adjustment commands. The pressure-release equalization control unit receives the three-dimensional spatial attitude change and the abnormal event determination result. When the abnormal event determination result is a tilting and jamming event, it applies circumferentially differentiated pressure-release compensation to subsequent pressure-release operations by adjusting the clamping force of the gripper corresponding to each circumferential position or the stroke difference of each pressure-release cylinder, so that the electrode axis is restored to the preset verticality range. When the abnormal event determination result is an overall slippage event, it outputs a slippage alarm signal and triggers the reset of the pressure-release adjustment command.

[0013] In one embodiment, the step of establishing a correlation model between electrode consumption rate and electrical parameters based on the pressure release sequence and electrical parameter sequence within a preset historical time window, and using the correlation model to predict the expected consumption rate of the electrode in real time based on the current electrical parameters, and generating a pressure release adjustment command, includes: Obtain the discharge amount data and timestamps of each real discharge event recorded by the electrode discharge measurement device within a preset historical time window, and form a discharge amount time series; Acquire the electrode current sequence, furnace power sequence, and electrode-to-ground voltage sequence collected by the smelting electrical parameter acquisition unit within the preset historical time window, and align the electrical parameter sequence with the pressure release time sequence according to the timestamp. The actual electrode consumption rate in each time period is calculated based on the pressure release time series. The actual electrode consumption rate is used as the output label, and the average electrode current, average furnace power, and average electrode-to-ground voltage in the corresponding time period are used as input features to construct a training sample set. The preset regression model is trained using the training sample set to obtain the correlation model between the electrode consumption rate and electrical parameters. The regression model includes any one of multiple linear regression model, support vector regression model, or neural network regression model. The system acquires the real-time electrode current, real-time furnace power, and real-time electrode-to-ground voltage at the current moment. It then inputs these data into the trained correlation model and outputs the expected consumption rate of the electrode. Based on the expected consumption rate and the preset pressure release cycle, the expected pressure release amount required for the next pressure release cycle is calculated, and a corresponding pressure release adjustment command is generated.

[0014] In one embodiment, the step of applying circumferentially differentiated pressure-release compensation to subsequent pressure-release operations by adjusting the clamping force of the gripper corresponding to each circumferential position or the stroke difference of each pressure-release cylinder, so as to restore the electrode axis to a preset perpendicularity range, includes:

[0015] Receive the current three-dimensional spatial attitude change output by the electrode pressing and releasing measurement device, and obtain the current tilt direction and current tilt angle of the electrode axis; Determine whether the current tilt angle exceeds the preset verticality range. If it does, determine the circumferential position corresponding to the tilt direction based on the tilt direction. At the circumferential position corresponding to the tilting orientation, increase the preset increment value of the clamping force of the corresponding gripper, or decrease the preset reduction value of the stroke of the corresponding pressure-release cylinder to form a circumferentially asymmetrical clamping force distribution or pressure-release stroke distribution. At the relative circumferential position of the tilted orientation, reduce the preset reduction value of the clamping force of the corresponding gripper, or increase the preset increment value of the stroke of the corresponding pressure-release cylinder. Write the circumferentially differentiated clamping force adjustment amount and the pressure release cylinder stroke adjustment amount into the execution command of the next pressure release operation so as to apply circumferentially differentiated pressure release compensation during the next pressure release. After the next pressing and releasing operation is completed, the three-dimensional spatial attitude change output by the electrode pressing and releasing measurement device is acquired again, and it is determined whether the tilt angle of the electrode axis has recovered to the preset verticality range. If it has not recovered, the circumferential differential pressing and releasing compensation is iteratively executed until the electrode axis is recovered to the preset verticality range.

[0016] Compared with the prior art, the beneficial effects of the present invention are: By simultaneously acquiring displacement signals relative to the electrode holder at each position through at least three non-contact displacement sensors distributed circumferentially around the electrode shell, the tilt direction and angle of the electrode axis during the pressing and releasing process are calculated, providing quantitative basis for attitude monitoring and anomaly diagnosis. This overcomes the limitation of traditional single-point measurement in sensing the spatial attitude of the electrode. Within a preset pressing and releasing time window, a true pressing and releasing event is determined only when all circumferential displacement signals monotonically increase and the total displacement falls within a preset range. This effectively filters out false triggers caused by vibration, signal noise, or local micro-movements, ensuring the accuracy and reliability of the pressing and releasing volume recording. Subsequently, a step back outside the window is identified as an overall slippage event, while a total displacement exceeding the upper limit within the window, combined with three-dimensional attitude recognition, is identified as a tilting and jamming event, enabling rapid attribution of anomaly types. Furthermore, by establishing a regression model between the consumption rate and electrical parameters based on historical pressing and releasing sequences and electrode current, power, and voltage sequences, the expected consumption rate is predicted in real time, and pressing and releasing adjustment commands are generated, achieving an intelligent upgrade from fixed-cycle pressing and releasing to on-demand pressing and releasing. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of an electrode pressure discharging measurement device in one embodiment; Figure 2 This is a structural block diagram of an intelligent smelting system for an electric arc furnace in one embodiment; Figure 3 This is a top view of the sensor's circumferential mounting position in one embodiment; Figure 4 This is a schematic diagram illustrating the establishment of a coordinate system and the transformation between radial distance and spatial point coordinates in one embodiment; Figure 5 This is a schematic diagram illustrating the definition of the line connecting the centers of two cross sections and the tilt angle / tilt direction in one embodiment. In the diagram: 102, non-contact displacement sensor; 104, data processing unit; 202, smelting electrical parameter acquisition unit; 204, smelting data processing unit; 206, pressure release equalization control unit. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0020] In one exemplary embodiment, such as Figure 1 , 3 As shown in Figure 5, an electrode pressing and discharging measuring device is provided, comprising: At least three non-contact displacement sensors 102 are circumferentially distributed around the same electrode shell. Each non-contact displacement sensor is fixedly installed at a different circumferential position of the electrode holder to synchronously acquire the displacement signal of the electrode shell relative to the electrode holder at the corresponding circumferential position.

[0021] Non-contact displacement sensors are sensing elements that can measure relative displacement without contacting the surface of the object being measured. Common types include eddy current displacement sensors, laser triangulation displacement sensors, or capacitive displacement sensors. They are suitable for harsh working conditions such as high temperature, vibration, and uneven surface of the electrode shell. The electrode shell is a cylindrical metal shell used in smelting equipment such as electric arc furnaces to protect the electrodes and transmit pressure. The electrode holder is a ring-shaped mechanical device fixed to the structure of the smelting equipment, surrounding the electrode shell, and used to guide and hold the electrodes. Circumferential distribution means that at least three non-contact displacement sensors are arranged around the circumference of the electrode shell at preset angular intervals. Each sensor is located on the same cross-section or different cross-sections, and its measurement direction points to the outer wall of the electrode shell. Synchronous acquisition means that each sensor simultaneously acquires the displacement signal of the corresponding position under the same trigger signal to truly reflect the instantaneous spatial pose of the electrode shell in the measurement cross-section.

[0022] For example, at two different heights above and below the electrode holder of the electric arc furnace, a set of eddy current displacement sensors is installed. Each set contains three sensors, arranged at three angular positions on the outer circumference of the electrode shell to form full circumferential coverage of the radial displacement of the electrode shell. Each sensor is fixed to the holder body by an L-shaped bracket and an insulating pad, with the probe end face maintaining a preset distance from the outer wall of the electrode shell to ensure that the sensor operates within its linear range. The signal lines of each sensor are connected to a synchronous data acquisition module. When the controller sends a synchronous trigger pulse, the displacement values ​​of the three measuring points are acquired at the same time reference. After analog-to-digital conversion, the values ​​are transmitted to the host computer via a fieldbus for real-time calculation of the radial displacement, skew direction, and axial compression of the electrode shell.

[0023] The data processing unit 104 receives the displacement signals collected by each non-contact displacement sensor and performs the following processing: The three-dimensional spatial attitude change of the electrode during the pressing and releasing process is calculated based on displacement signals at at least three circumferential positions. The three-dimensional spatial attitude change includes at least the tilt direction and tilt angle of the electrode axis.

[0024] Among them, the displacement signals at at least three circumferential positions refer to the radial displacement values ​​collected at the same time by multiple non-contact displacement sensors distributed around the circumference of the electrode shell. Each sensor measures the change in distance between the outer wall of the electrode shell at its location and the gripper. The three-dimensional spatial attitude change refers to the change in the orientation of the geometric axis of the electrode shell in space during the pressing and releasing process, including two main parameters: the direction angle of the axis tilt and the tilt angle. The tilt direction of the electrode axis refers to the azimuth angle pointed to by the projection of the electrode axis onto the horizontal plane after it deviates from the vertical direction. The tilt angle is the angle between the electrode axis and the vertical direction, which characterizes the degree to which the electrode deviates from the vertical state.

[0025] For example, after receiving real-time data from three synchronously acquired displacement sensors, the host computer first converts the voltage or current signals into actual displacement values ​​according to the calibration parameters of each sensor. Then, based on the known installation azimuth angles of the three sensors on the circumference, the displacement value of each measuring point is decomposed into components along two orthogonal horizontal axes. Through geometric synthesis, the center offset vector of the electrode shell at the measuring section is calculated. The magnitude of this vector is the eccentricity of the section, and its direction is the eccentricity direction of the section. If a set of sensors is arranged at different heights above and below the electrode holder, the center coordinates of the upper and lower sections can be calculated respectively. The direction of the line connecting the centers of the two sections is taken as the current actual axis direction of the electrode. By comparing this axis direction with the vertical direction, the tilt direction and tilt angle of the axis can be obtained. By continuously performing the above calculations during the pressing and releasing process, the continuous change process of the electrode's three-dimensional spatial attitude can be obtained in real time.

[0026] For example, the installation angle of each non-contact displacement sensor in the circumferential direction of the electrode shell and the corresponding real-time displacement value are obtained; with the electrode holder as a reference, the spatial discrete point coordinates of the bottom end face of the electrode shell at each sampling time are constructed according to the installation angle and real-time displacement value of each sensor; the spatial discrete point coordinates are fitted to obtain the spatial plane equation of the bottom end face of the electrode shell, and the normal vector of the spatial plane is calculated; the normal vector is compared with the gravity direction vector, and the angle of the normal vector deviating from the gravity direction is determined as the tilt angle of the electrode axis, and the projection direction of the normal vector on the horizontal plane is taken as the tilt direction.

[0027] In this embodiment, please refer to the specific details. Figure 3 The diagram shows the electrode housing, electrode holder, and three non-contact displacement sensors circumferentially mounted at 0°, 120°, and 240° positions, along with their radial measurement directions pointing towards the electrode housing axis. Three eddy current displacement sensors are respectively installed at 0°, 120°, and 240° positions around the circumference of the electrode holder. The system first reads the real-time displacement values ​​of each sensor and superimposes them with the initial installation gap of the sensor probes to obtain the radial distance from each measuring point on the outer wall of the electrode housing to the center of the holder. See details in the reference section. Figure 4 This diagram illustrates the conversion relationship between the spatial positions and radial distances of measuring points P1, P2, and P3 on the outer wall of the electrode housing to spatial point coordinates in a coordinate system O-XYZ with the axis of the gripper as the Z-axis. A fixed coordinate system is established with the electrode gripper as the reference reference: the central axis of the electrode gripper is taken as the Z-axis, and the reference cross-section (e.g., the sensor mounting section) is taken as the XOY plane, establishing a right-handed rectangular coordinate system O-XYZ. Let the center of the probe end face of the i-th non-contact displacement sensor be located at a point with the Z-axis as the center and a radius of... On the circumference ( The installation radius from the center of the probe end face to the reference axis of the holder is , and the circumferential installation azimuth angle is . Installation height is The sensor's measurement direction is radially aligned with the axis of the electrode housing, and the initial installation gap between the probe end face and the outer wall of the electrode housing is [missing information]. (The installation calibration time is known).

[0028] At any sampling time Let the change in relative displacement output by the sensor be... Then, the gap between the probe end face and the outer wall of the electrode shell at the current moment is: ; The measured point on the outer wall of the electrode shell Located on the ray in the direction of probe measurement, at a distance from the probe end face At this location, its three-dimensional coordinates in the reference coordinate system of the gripper are: ; in, This is the radial distance of the measuring point relative to the axis of the gripper in the horizontal plane. Determine its orientation in the horizontal plane. Its height coordinates. Therefore, a radial distance reading for each sensor, combined with its installation angle... and installation parameters This uniquely identifies a discrete spatial point on the outer wall of the electrode housing. At least three sensors at circumferential positions synchronously acquire data at the same sampling time, resulting in at least three discrete spatial points P1, P2, and P3. Combining the installation angles of each sensor, the planar coordinates of each measuring point in the grip cross-sectional coordinate system can be calculated. Then, each measuring point is assigned a height coordinate according to the sensor's installation height, yielding three discrete spatial points. Using these three points, a spatial plane fitting is performed to calculate the normal vector of the plane. See details for further information. Figure 5 The diagram shows the vertical direction k, the center of the lower section Od, the center of the upper section Ou, the direction of the electrode axis v, the tilt angle α, and the azimuth angle φ projected onto the horizontal plane by the tilt direction. When a set of sensors is placed at two different heights above and below the electrode holder, and the coordinates of the center of the lower section are Od(xd,yd,zd) and the coordinates of the center of the upper section are Ou(xu,yu,zu), then the vector of the electrode axis direction is... Let the unit vector in the vertical direction be... Then the tilt angle of the electrode axis is The tilt direction of the electrode axis is the projection direction of the axis direction vector onto the horizontal plane (XOY plane), and its azimuth angle is... .

[0029] The center of each cross-section is obtained by fitting the coordinates of at least three measuring points on the outer wall of the cross-section using a three-point circle or least squares circle fitting method: Let the coordinates of the three measuring points inside the cross-section be (x1, y1), (x2, y2), and (x3, y3), and substitute them into the circle equation. Solving the system of equations yields the coordinates of the circle's center (x0, y0) and its radius r. When using the plane fitting method, it is assumed that the discrete points in space satisfy the plane equations. If we use least squares to find the coefficients A, B, and C, and the normal vector n = (A, B, C), then the tilt angle is... The tilt direction is the azimuth angle of the projection of the normal vector onto the horizontal plane. The normal vector is compared with the gravity direction vector fixed to the device; the angle between the two vectors is the tilt angle of the electrode axis. The tilt direction is the angle between the projection of the normal vector onto the horizontal plane and a preset reference direction. This result is updated in real time and used for attitude monitoring during the pressing and releasing process.

[0030] Extract the variation characteristics of displacement signals at each circumferential position over time. If the displacement signals at all circumferential positions are detected to be monotonically increasing within the preset compression time window, and the total displacement at each circumferential position falls within the preset compression range, it is determined to be a real compression event and the compression amount for that time is recorded.

[0031] Among them, the change characteristics of the circumferential position displacement signal over time refer to the increasing and decreasing trend and fluctuation law of the displacement values ​​collected by each non-contact displacement sensor on the time axis; the preset pressure release time window is the maximum allowable duration of the pressure release action set according to the electrode pressure release process specification. Within this time window, the dynamic changes of the displacement signal are monitored to determine whether a real pressure release has occurred; monotonically increasing means that the displacement signal continuously increases within the pressure release time window without obvious contraction or fluctuation, indicating that the electrode shell is continuously moving downward; the preset pressure release amount range is the allowable displacement range of a single pressure release set according to the electrode consumption compensation requirements of the smelting process. The total displacement falling into this range indicates that the pressure release amount meets the process requirements; the real pressure release event refers to an effective pressure release action generated by the pressure release mechanism actively pushing down the electrode shell, which is different from the thermal expansion and contraction of the electrode shell, loosening of the holder, or sensor noise interference; the pressure release amount of this time is the average value of the total displacement of each circumferential position, used to reflect the overall downward movement distance of this pressure release.

[0032] For example, signals from three displacement sensors are continuously acquired at a fixed sampling period, and the displacement trend of each sensor within the current sliding time window is calculated in real time. When it is detected that the displacement values ​​of the three sensors all show a monotonically increasing characteristic within a continuous preset time window, and the cumulative displacement value of each sensor is within the preset release range at the end of the window, a real release event is determined to have occurred. The system records the average value of the cumulative displacement values ​​of the three sensors as the release amount for this event, saves it to the release history database, and marks the release timestamp. If the displacement trend of a certain sensor is inconsistent with that of other sensors, or the total displacement exceeds the preset range, it is determined to be a non-real release or abnormal release, an alarm message is generated, and the operator is prompted to check the electrode holder status and sensor working status.

[0033] For example, extracting the time-varying characteristics of displacement signals at each circumferential position includes: performing real-time sliding median filtering or mean filtering on the displacement signals at each circumferential position to remove high-frequency noise interference; performing first-order difference operation on the filtered displacement signals to obtain the instantaneous displacement change rate at each sampling moment; and summing the filtered displacement signals at each circumferential position according to a preset time interval or number of sampling points to obtain the total displacement of each circumferential position within any specified time period; wherein, the variation characteristics include a real-time displacement value sequence, an instantaneous displacement change rate sequence, and the total displacement within the specified time period.

[0034] In this embodiment, real-time sliding median filtering refers to taking a neighborhood window centered on the current sampling point and taking several points before and after it, replacing the current sampling value with the median of all sampling values ​​within the window, effectively suppressing isolated impulse noise while maintaining the signal step change characteristics; mean filtering replaces the current sampling value with the arithmetic mean of the sampling values ​​within the window, suitable for filtering out high-frequency random noise. First-order difference operation refers to the difference between the displacement values ​​of adjacent sampling times in the filtered displacement sequence to obtain the displacement increment within the sampling interval, which, divided by the sampling interval, is the instantaneous displacement change rate, reflecting the downward movement speed of the electrode shell at the corresponding circumferential position. The total displacement refers to the algebraic sum of the displacement increments of all sampling intervals at each circumferential position within a specified time period, representing the cumulative downward movement distance of the electrode shell relative to the gripper during this period. The change characteristics include the real-time displacement value sequence, the instantaneous displacement change rate sequence, and the total displacement within the specified time period, which together constitute the data basis for determining the press-release event; signals from three displacement sensors are acquired at a fixed sampling frequency, and a sliding window median filter with a width of five sampling points is applied to the data stream of each sensor to remove instantaneous pulses caused by electromagnetic interference or surface corrosion. The filtered displacement sequence is processed point-by-point by calculating backward difference to obtain the instantaneous downward displacement rate per second. Simultaneously, the system maintains a preset-length sliding accumulation window, continuously accumulating the filtered displacement increments and outputting the total cumulative displacement over the most recent period in real time. During the compression and release process, the system monitored that the total cumulative displacement of all three sensors continuously increased, and the instantaneous rate was significantly higher than the background fluctuation level during the compression and release mechanism's operation. Based on this, the complete change characteristics of this compression and release event are extracted for the determination of the actual compression and release event and the recording of the compression and release amount.

[0035] For example, when all displacement signals at all circumferential positions are detected to be monotonically increasing within a preset deflation time window, and the total displacement at each circumferential position falls within a preset deflation range, it is determined to be a real deflation event and the deflation amount is recorded. This includes: taking the moment when the instantaneous displacement change rate at at least one circumferential position is first detected to exceed a preset start threshold as the start time of the deflation time window, and using a preset fixed duration as the length of the deflation time window; within the deflation time window, continuously determining whether the difference between the displacement signals at each circumferential position at adjacent sampling moments is greater than or equal to zero, if any channel If a negative difference occurs, the current judgment is terminated. At the end of the compression time window, the displacement difference between the end and start times of each circumferential position is calculated as the total displacement of that circumferential position. The total displacement of each circumferential position is compared with the lower and upper limits of the preset compression range. If the displacement signals of all circumferential positions within the window are monotonically increasing and the total displacement of each circumferential position falls within the preset compression range, it is determined to be a real compression event, and the average or weighted average of the total displacement of all circumferential positions is recorded as the compression amount.

[0036] In this embodiment, the preset start threshold is a threshold value set for the instantaneous displacement change rate, used to distinguish between rapid displacement caused by the active downward push of the pressing and releasing mechanism and slow changes such as temperature drift; the start time of the pressing and releasing time window is based on the first detection that the instantaneous rate of any circumferential position exceeds the threshold, and the window length is preset to a fixed value based on the typical duration of the pressing and releasing mechanism's action; monotonically increasing means that the increment of the displacement value at each sampling time within the window relative to the previous time is non-negative, to exclude false displacement caused by loosening or rebound of the gripper; the total displacement is taken as the difference between the displacement value at the end time and the start time of the window; the preset pressing and releasing amount range is given by the process specification, including a lower limit and an upper limit, used to prevent incomplete pressing and over-pressing and releasing, respectively; when all circumferential positions simultaneously satisfy monotonically increasing and the total displacement falls within the range, it is determined to be a real pressing and releasing event, and the pressing and releasing amount can be taken as the arithmetic mean of the total displacement of each circumferential direction or the weighted average based on the sensor confidence level; a real-time rate calculation thread is maintained for each displacement sensor channel. When the instantaneous rate of a channel first exceeds the preset start threshold, the compression and release judgment state machine is triggered to enter the window monitoring state, and a window timer is started simultaneously. During the window duration, for each frame of data acquired, the difference between the current displacement of each channel and the previous sampling point is immediately calculated. If any channel has a negative difference value, the state machine immediately resets and exits the current judgment. After the window timer expires, the displacement values ​​of each channel at the end and beginning of the window are read and the difference is calculated to obtain the total displacement of each channel. These total displacements are compared one by one with the preset compression and release range. If all of them fall within the range, it is judged as a real compression and release event. The system writes the average of the total displacement of each channel into the compression and release record database and marks it with a timestamp and event number. If the total displacement of any channel exceeds the range, an abnormal alarm is generated, prompting an inspection of the compression and release mechanism.

[0037] When a step back is detected in the displacement signal of at least one circumferential position outside the preset compression time window, or when the total displacement of the displacement signal of at least one circumferential position within the preset compression time window exceeds the upper limit of the preset compression range, the abnormal event that has occurred is determined to be an overall slippage event or a tilting and jamming event based on the three-dimensional spatial attitude change.

[0038] Among them, the step back refers to a large negative jump in the displacement signal between adjacent sampling times during a period outside the preset pressure release time window. That is, the electrode shell moves upward relative to the gripper when there is no pressure release command, and its rate of change is significantly higher than the slow fluctuation caused by normal thermal expansion and contraction. The overall slip event refers to insufficient clamping force between the electrode shell and the gripper, which causes the electrode shell to slide downward or retreat upward uniformly under the action of gravity or external force. At this time, the displacement signal change amplitude of each circumferential position is similar, and the three-dimensional spatial attitude change remains basically unchanged. The tilt jamming event refers to the electrode shell being subjected to uneven local force or obstruction by foreign objects, resulting in normal displacement on one side and obstructed or reversed displacement on the other side. At this time, the displacement signal of each circumferential position shows significant asymmetry, and the three-dimensional spatial attitude change is manifested as a sudden increase in tilt angle.

[0039] For example, the system continuously monitors the signals from each displacement sensor. If, at a certain moment, a significant negative jump in the displacement value of one or more channels is detected between adjacent sampling points, and this jump occurs outside any known release time window, and the jump amplitude exceeds a preset step back threshold, an abnormal back-off event is determined to have occurred. The system then reads the current three-dimensional spatial attitude change: if the back-off amounts of each circumferential channel are similar, and the tilt angle has not changed significantly relative to the value at the end of the release, it is determined to be an overall slippage event; if the back-off is mainly concentrated on one side of the sensor, and the tilt angle data shows a significant increase, it is determined to be a tilt jamming event. Based on the determination results, the system generates corresponding alarm information: an overall slippage alarm prompts a check of the gripper clamping pressure, and a tilt jamming alarm prompts a check of the electrode shell surface condition and guide clearance.

[0040] For example, detecting a step back in the displacement signal of at least one circumferential position outside a preset deflation time window includes: real-time monitoring of the displacement signal of each circumferential position; when it is detected that the current moment is not within any triggered deflation time window, calculating the displacement difference between the current moment and the previous moment for each circumferential position; if the current displacement difference of a certain circumferential position is negative and its absolute value exceeds a preset step back threshold, and the displacement signal continues to decrease and the total cumulative decrease exceeds a preset cumulative back threshold in multiple consecutive sampling moments, then it is determined that a step back has occurred at the circumferential position, and the event and the circumferential position where it occurred are recorded.

[0041] In this embodiment, "the current moment is not within any triggered release time window" means that the current state of the system is neither the duration of the release window nor the instant the window has just been triggered. Only backtracking that occurs outside the window is considered abnormal backtracking, in order to avoid misjudging the normal rebound after the release is completed as abnormal. The displacement difference refers to the difference between the displacement values ​​of two adjacent sampling moments. A negative value indicates that the electrode shell is moving away from the sensor, i.e., backtracking has occurred. The step backtracking threshold is the lower limit of the absolute value of the displacement difference set to distinguish between normal small fluctuations and abnormal large backtracking. This threshold is determined based on the statistical fluctuation range of normal thermal expansion and contraction and mechanical vibration of the electrode shell. The continuous decrease of the displacement signal within multiple consecutive sampling moments means that the backtracking is not a single-point jump, but has a certain continuity in time, in order to eliminate the interference of isolated pulse noise. The cumulative backtracking threshold is a constraint on the total backtracking displacement. Only when the cumulative backtracking exceeds this threshold is a step backtracking finally confirmed. Backtracking detection is performed on each circumferential displacement channel in each sampling cycle. First, check if the current time is within any compression / release time window. If so, skip the backtracking detection for that channel. Otherwise, calculate the difference between the current displacement value and the displacement value at the previous sampling time. When this difference is negative and its absolute value is greater than a preset step backtracking threshold, the system marks the channel as a backtracking candidate and begins accumulating the backtracking amount. Subsequently, the system continues to monitor the displacement value of the channel at subsequent consecutive sampling times. As long as the displacement continues to decrease, the absolute value of each negative difference is added to the backtracking accumulation. When the backtracking accumulation exceeds a preset accumulated backtracking threshold, a step backtracking event is determined to have occurred for that channel, and the event time, channel number, and accumulated backtracking amount are recorded. If the displacement rebounds during the accumulation process and the rebound exceeds the fluctuation tolerance, the backtracking candidate status is canceled, and the accumulated amount is reset.

[0042] For example, detecting that the total displacement of at least one circumferential position exceeds the upper limit of a preset release range within a preset release time window includes: accumulating the displacement of each circumferential position in real time within each release time window, and comparing the real-time accumulated displacement with the upper limit of the preset release range; when the real-time accumulated displacement of any circumferential position exceeds the upper limit of the preset release range, immediately triggering an over-range flag, and recording the time corresponding to triggering the over-range flag and the information of the corresponding circumferential position, while continuing monitoring until the end of the window.

[0043] In this embodiment, within the compression time window, for each frame of displacement data acquired, the system calculates the difference between the current displacement and the initial displacement of the window for each circumferential channel as the real-time cumulative displacement. The system compares this value with the upper limit of the preset compression range. Once the cumulative displacement of any channel exceeds this upper limit, the system immediately sets the over-range flag for that channel, records the current time, channel number, and current cumulative displacement, and pushes an over-range warning to the monitoring backend. During the remaining time of the window, the system continues to record the displacement data of each channel, but does not remove the set flag even if the cumulative displacement subsequently decreases. All over-range events are summarized at the end of the window to determine the anomaly type.

[0044] For example, based on the change in three-dimensional spatial attitude, determining whether the current abnormal event is an overall slippage event or a tilting and jamming event includes: when a step back event or a total displacement exceeding the limit event is detected, obtaining the tilt angle and tilt direction of the electrode axis at the current moment; comparing the tilt angle with a preset tilt threshold; if the tilt angle is less than the tilt threshold, and the displacement signals at each circumferential position change in the same direction and with similar magnitudes, then determining the current abnormal event as an overall slippage event; if the tilt angle is greater than or equal to the tilt threshold, and there are at least two circumferential positions with opposite trends in displacement signal changes or displacement differences exceeding a preset inconsistency threshold, then determining the current abnormal event as a tilting and jamming event.

[0045] In this embodiment, when the system detects that a channel experiences a step back outside the window or that the cumulative displacement of a channel exceeds the limit, an anomaly judgment process is immediately triggered. The system first reads the current three-dimensional spatial attitude calculation results to obtain the tilt angle and tilt direction of the electrode axis. If the tilt angle is less than a preset tilt threshold, and the displacement changes of the three sensors are all positive or all negative, and the difference between the maximum and minimum values ​​of the three changes is less than the inconsistency threshold, it is judged as an overall slippage event, generating an "overall slippage" alarm, prompting the user to check the clamping pressure of the gripper and the surface condition of the electrode shell. If the tilt angle is greater than or equal to the tilt threshold, and the displacement change direction of one sensor is opposite to that of the other two, or the difference in displacement change amplitude between any two sensors exceeds the inconsistency threshold, it is judged as a tilt jamming event, generating a "tilt jamming" alarm, marking the tilt direction and the channel with the maximum displacement difference in the alarm information, prompting the user to check the guide gap and the surface of the electrode shell in the corresponding direction.

[0046] For example, let the tilt threshold be... Inconsistency threshold is and boundary tolerance When a step rollback event or a total displacement exceeding the limit event is detected, the anomaly type shall be determined and handled according to the following rules: If the tilt angle Furthermore, the displacement signals at each circumferential position change in the same direction, and the difference in displacement amplitude between any two circumferential positions is ≤ If the event is detected as a slippage event, a slippage alarm signal will be output, and the reset of the press-release adjustment command will be triggered. If the tilt angle However, there are at least two circumferential positions where the displacement signals change in different directions, or the difference in displacement amplitude is greater than 1. If the event is detected as a suspected local anomaly (a precursor to tilting and jamming): record the event and output an early warning, while continuously monitoring the tilt angle and displacement difference of the subsequent release window. If the tilt angle rises to ≥ If the displacement difference continues to increase, the event will be upgraded to a tilting and jamming event and corresponding measures will be taken. If the tilt angle Furthermore, there are at least two circumferential positions where the displacement signals show opposite trends, or the difference in displacement amplitude is greater than 1. If the event is identified as a tilting and jamming event, a tilting and jamming alarm will be output (marking the tilting direction and the circumferential position with the greatest displacement difference), and circumferential differential pressure release compensation correction will be performed. If the tilt angle However, the displacement signals at each circumferential position change in the same direction and the difference in their amplitudes is ≤ If the overall slippage is accompanied by tilt, it is determined to be an overall slippage event (output slippage alarm and reset the press-release adjustment command), and a tilt correction prompt is attached to the alarm information. If the difference between the tilt angle and the tilt threshold, or the difference between the displacement change amplitude and the inconsistency threshold, falls within the boundary tolerance zone. within (i.e. or If the condition is found to be abnormal, it should be handled strictly as a tilting and jamming event, and the original sampling data should be retained for manual review. If the current tilt angle calculation result is missing or the sensor data is invalid, it is determined to be a data anomaly: output a data anomaly alarm and suspend the execution of this pressing and releasing command; In addition to the above situations, other compression and release abnormalities are identified: a general abnormality alarm is output, the execution of subsequent compression and release commands is suspended, and resumption is only possible after manual confirmation.

[0047] The above judgment rules cover all combinations of judgment elements. Among them, the tilt threshold... Inconsistency threshold and boundary tolerance The value can be determined based on the statistical distribution of historical normal pressure release data and calibration tests (for example, by taking the mean of the normal operating condition statistics plus k times the standard deviation), and adjusted during the on-site commissioning stage.

[0048] Based on the same inventive concept, this application also provides an intelligent smelting system for a submerged arc furnace to implement the electrode pressure release measuring device described above. The solution provided by this system is similar to the solution described in the above method. Therefore, the specific limitations of one or more embodiments of the intelligent smelting system for a submerged arc furnace provided below can be found in the limitations of the electrode pressure release measuring device described above, and will not be repeated here.

[0049] In one exemplary embodiment, such as Figure 2 As shown, an intelligent smelting system for an electric arc furnace is provided, comprising: The electrode pressure measurement device described above.

[0050] The smelting electrical parameter acquisition unit 202 is used to acquire electrode current, furnace power and electrode-to-ground voltage in real time.

[0051] The smelting electrical parameter acquisition unit is a measurement module used to acquire key electrical parameters of the smelting process in the submerged arc furnace. Electrode current refers to the large alternating current flowing through each electrode, measured non-contactly by a current transformer or Rogowski coil installed on the electrode short grid or conductive copper busbar. Furnace power includes active power, reactive power, and apparent power, obtained through instantaneous power calculation and integration of the three-phase voltage and current signals. Electrode-to-ground voltage refers to the voltage at the electrode tip relative to the furnace bottom grounding point, measured using a high-voltage isolation transformer and voltage divider. These electrical parameters, together with the displacement and attitude data output by the electrode pressure measurement device, constitute multi-dimensional information about the smelting process, used for real-time furnace condition assessment and electrode adjustment guidance.

[0052] For example, the smelting electrical parameter acquisition unit consists of three sets of Rogowski coil current transformers, three voltage transformers, and a multi-functional electrical parameter acquisition module. The Rogowski coils are mounted on the conductive copper tubes of each phase electrode, and their output signals are converted into voltage signals proportional to the current by an integrator. The high-voltage terminals of the voltage transformers are connected to the electrode holders, and the low-voltage terminals are connected to the voltage input channel of the acquisition module. The acquisition module synchronously samples the three-phase current and voltage at a rate of one hundred times per second, calculates the effective values ​​of the fundamental wave and each harmonic through fast Fourier transform, and simultaneously calculates the total three-phase power in real time. The calculation results are uploaded to the process controller via the Modbus TCP protocol and used in conjunction with the electrode displacement data for adaptive control of the electrode insertion depth.

[0053] The smelting data processing unit 204 receives the pressure release data, three-dimensional spatial attitude change and abnormal event judgment results output by the electrode pressure release measurement device, and receives the electrical parameter data output by the smelting electrical parameter acquisition unit.

[0054] The smelting data processing unit is the core computing module of the intelligent smelting system for electric arc furnaces. It is responsible for establishing a quantitative correlation between electrode consumption and electrical parameters and generating control commands. The preset historical time window refers to the data selection period used for modeling, usually taking continuous smelting data from the most recent few hours to several days to reflect the current furnace condition and electrode status. The pressure release sequence refers to the time series of pressure release amounts of each real pressure release event recorded by the electrode pressure release measurement device within the time window. The cumulative pressure release amount is the total consumption of the electrode during that period. The electrical parameter sequence refers to the time series of electrical parameters such as electrode current and furnace power recorded by the smelting electrical parameter acquisition unit within the same time window. The correlation model is a mathematical model established through regression analysis or machine learning methods, with electrical parameters as input and electrode consumption rate as output. The expected consumption rate refers to the expected consumption length of the electrode per unit time predicted by the correlation model based on the current real-time electrode current, real-time furnace power, and real-time electrode-to-ground voltage. The pressure release adjustment command is a pressure release control signal generated by the smelting data processing unit based on the predicted consumption rate and the current electrode position status, used to adjust the timing and amount of electrode pressure release.

[0055] For example, the smelting data processing unit is deployed on an industrial server. It extracts pressure release records and electrical parameter records from the historical database for the past few hours to construct a training dataset. Using multiple linear regression or support vector regression methods, with the mean electrode current and mean furnace active power as input features and the average electrode consumption rate within the corresponding time period as the output label, a correlation model is trained. During online operation, the data processing unit reads the mean electrical parameters of the sliding window for the previous 30 minutes at preset intervals and inputs them into the correlation model to obtain the predicted consumption rate. This predicted consumption rate is compared with the current actual electrode position and the last pressure release time fed back by the electrode pressure release measurement device. If the required pressure release interval corresponding to the predicted consumption rate has been reached, a pressure release adjustment command is generated, including a suggested pressure release amount, and sent to the process controller to execute the pressure release operation.

[0056] For example, the smelting data processing unit establishes a correlation model between the electrode consumption rate and the electrical parameters based on the pressure release sequence and electrical parameter sequence within a preset historical time window. Using the correlation model, it predicts the expected consumption rate of the electrode in real time based on the current electrical parameters and generates a pressure release adjustment command.

[0057] Furthermore, based on the pressure release sequence and electrical parameter sequence within a preset historical time window, a correlation model between electrode consumption rate and electrical parameters is established. This correlation model is used to predict the expected electrode consumption rate in real time based on the current electrical parameters, and to generate pressure release adjustment commands. This includes: acquiring pressure release data and timestamps of all real pressure release events recorded by the electrode pressure release measurement device within the preset historical time window, forming a pressure release time series; acquiring the electrode current sequence, furnace power sequence, and electrode-to-ground voltage sequence collected by the smelting electrical parameter acquisition unit within the preset historical time window, and aligning the electrical parameter sequence with the pressure release time series by timestamp; calculating the actual electrode consumption rate within each time period based on the pressure release time series, and using the actual electrode consumption rate as the output label. The system constructs a training sample set by using the average electrode current, average furnace power, and average electrode-to-ground voltage over a corresponding time period as input features. It then trains a pre-defined regression model using this training sample set to obtain a correlation model between the electrode consumption rate and the electrical parameters. The regression model can be any one of multiple linear regression, support vector regression, or neural network regression. The system acquires the real-time electrode current, real-time furnace power, and real-time electrode-to-ground voltage at the current moment, inputs these values ​​into the trained correlation model, and outputs the expected electrode consumption rate. Based on the expected consumption rate and the pre-defined pressure release cycle, it calculates the expected pressure release amount required for the next pressure release cycle and generates the corresponding pressure release adjustment command.

[0058] In this embodiment, the preset historical time window refers to a continuous historical smelting period used to train the correlation model. Its length is determined based on the stability of the furnace condition and the statistical law of electrode consumption, and usually covers multiple pressure release cycles to ensure sample representativeness. The pressure release volume sequence is a time series composed of the pressure release volume and timestamp of each real pressure release event recorded by the electrode pressure release measurement device. The relationship between the cumulative pressure release volume and time directly reflects the actual consumption of the electrode. The electrical parameter sequence includes electrode current, furnace power, and electrode-to-ground voltage. These electrical parameters are closely related to arc heat, molten pool temperature, and electrode burn-off rate. The correlation model is a quantitative mapping between electrical parameters and electrode consumption rate established through regression analysis, so that the consumption rate can be inferred from the electrical parameters when there is no direct measurement. The regression model can be selected from multiple linear regression, support vector regression, or neural network regression according to the process complexity and data characteristics.

[0059] For example, when the correlation model adopts a neural network regression model, the network structure and specific training method of the correlation model are as follows: the number of nodes in the input layer is equal to the dimension of the input features (mean electrode current, mean furnace power, mean electrode voltage to ground, a total of 3), the hidden layer adopts a 1-2 layer fully connected structure (8-16 nodes per layer, ReLU activation function), the output layer has 1 node, and the output electrode is the expected consumption rate; the loss function adopts mean squared error (MSE), and the optimizer adopts Adam or stochastic gradient descent; the network size is matched with the training sample size, and an early stopping strategy is adopted during training to prevent overfitting.

[0060] The pressure release equalization control unit 206 receives three-dimensional spatial attitude changes and abnormal event judgment results. When the abnormal event judgment result is a tilting and jamming event, it applies circumferential differential pressure release compensation to subsequent pressure release operations by adjusting the clamping force of the gripper corresponding to each circumferential position or the stroke difference of each pressure release cylinder, so that the electrode axis is restored to the preset verticality range. When the abnormal event judgment result is an overall slippage event, it outputs a slippage alarm signal and triggers the reset of the pressure release adjustment command.

[0061] The pressure release equalization control unit is an execution module that receives the three-dimensional spatial attitude change and abnormal event judgment results output by the electrode pressure release measuring device, and performs circumferential differential compensation control on the electrode pressure release process. The tilt jamming event refers to the abnormal working condition in which the movement of one side of the electrode shell is blocked due to local deformation or uneven guide clearance, and the tilt angle of the electrode axis exceeds the preset tilt threshold. The circumferential differential pressure release compensation refers to adjusting the clamping force of the gripper or the stroke of the pressure release cylinder corresponding to each circumferential position independently so that the pressure release amount on the tilted side is greater or less than that on the opposite side, thereby generating a corrective torque in the subsequent pressure release process and gradually restoring the electrode axis to the preset verticality range. The overall slippage event refers to the abnormality caused by the electrode shell sliding down or retracting uniformly as a whole, and the clamping force is insufficient. In this case, no differential compensation is performed, but a slippage alarm signal is issued and the smelting data processing unit is instructed to re-evaluate the pressure release amount and reset the current pressure release adjustment command.

[0062] For example, the pressure release equalization control unit consists of a PID controller in a PLC and a hydraulic proportional valve group. When the received abnormal event determination result is a tilting jamming event, the controller reads the electrode tilt direction and tilt angle to determine the circumferential position corresponding to the tilt. Then, by reducing the clamping force of the holder on the tilted side or increasing the stroke of the pressure release cylinder on that side, the electrode shell on that side obtains a larger downward movement during the next pressure release, while the opposite side maintains the original parameters or makes a reverse fine adjustment, thereby generating a reverse bending moment to correct the tilt. After each pressure release, the controller recalculates the tilt angle based on the new attitude data. If the tilt angle still does not recover to the preset verticality range, compensation is continuously applied in the next pressure release cycle until the tilt angle converges. When the determination result is an overall slippage event, the controller immediately sets the slippage alarm output, cuts off the execution of the current pressure release command, and sends a "pressure release failure" signal to the smelting data processing unit. The data processing unit then recalculates the required pressure release amount based on the latest electrical parameters and the current electrode position, generating a corrected pressure release adjustment command.

[0063] For example, by adjusting the clamping force of the gripper corresponding to each circumferential position or the stroke difference of each pressure-release cylinder, circumferentially differentiated pressure-release compensation is applied to subsequent pressure-release operations to restore the electrode axis to a preset perpendicularity range. This includes: receiving the current three-dimensional spatial attitude change output by the electrode pressure-release measuring device, obtaining the current tilt direction and current tilt angle of the electrode axis; determining whether the current tilt angle exceeds the preset perpendicularity range; if it does, determining the circumferential position corresponding to the tilt orientation based on the tilt direction; at the circumferential position corresponding to the tilt orientation, increasing the preset increment value of the clamping force of the corresponding gripper, or decreasing the preset reduction value of the stroke of the corresponding pressure-release cylinder, to form a circumferentially asymmetrical clamping force. The distribution or compression stroke distribution; at the relative circumferential position of the tilt direction, reduce the preset reduction value of the clamping force of the corresponding gripper, or increase the preset increment value of the stroke of the corresponding compression cylinder; write the circumferentially differentiated gripper clamping force adjustment amount and compression cylinder stroke adjustment amount into the execution instruction of the next compression operation, so as to apply circumferentially differentiated compression compensation during the next compression; after the next compression operation is completed, re-acquire the three-dimensional spatial attitude change output by the electrode compression measurement device, and determine whether the tilt angle of the electrode axis has recovered to the preset verticality range. If it has not recovered, continue to iteratively execute the circumferentially differentiated compression compensation until the electrode axis recovers to the preset verticality range.

[0064] In this embodiment, after receiving a signal that the electrode tilt angle exceeds a preset verticality range, the pressure-release equalization control unit determines the tilt direction to be due east, corresponding to the sensor position on the easternmost side of the outer wall of the electrode housing. In subsequent pressure-release commands, for the circumferential position at this tilt direction, the clamping pressure value of the gripper is increased by a preset increment, while the target stroke of the pressure-release cylinder corresponding to that position is shortened by a preset reduction, thereby relatively reducing the downward movement of that side during pressure-release; at the opposite western position, the clamping pressure is decreased by a preset reduction, and the cylinder stroke is increased by a preset increment, thereby relatively increasing the downward movement of that side. The above differentiated settings are written into the next pressure-release command. After executing the pressure-release, the electrode posture is checked again. If the tilt angle still has not returned to the allowable range, iterative adjustments continue until the axis returns to verticality.

[0065] For example, the preset increment / decrement value is determined using a closed-loop proportional adjustment method based on the tilt angle deviation. Let the current tilt angle be... The preset verticality target angle is (Usually 0° or the upper limit of the allowable verticality is taken), then the adjustment amount for this operation is as follows: Confirmed. Among them, The proportional coefficient and the control coefficient are determined by on-site calibration tests based on the characteristics of the controlled object (electrode weight, the transmission relationship between clamping force and release amount, and the proportional relationship between cylinder stroke and release amount). The adjustment amount Δ has an upper limit (e.g., not exceeding 5% to 10% of a single release stroke) to prevent over-adjustment or system instability caused by excessive adjustment in one step. A larger step size can be used when the deviation is large, and a smaller step size can be used when the deviation is small, in order to speed up convergence and improve steady-state accuracy.

[0066] Iterative convergence condition: After each compensated pressing and releasing operation, the tilt angle of the electrode axis is reacquired; convergence occurs when the current tilt angle α does not exceed the upper limit of the preset perpendicularity range, and the change in tilt angle after two adjacent pressing and releasing operations is met. If the convergence threshold is less than the preset threshold, the correction is considered to have converged, and compensation is stopped. If convergence is not achieved after the preset maximum number of iterations, a correction failure alarm is output and manual processing is initiated. If necessary, a damping coefficient is introduced to reduce the compensation amount with each iteration, thus preventing compensation oscillations.

[0067] The selection criteria for adjustment methods are as follows: Clamping force adjustment is suitable for electrode holding mechanisms with controllable hydraulic clamping force, where the main cause of abnormalities is uneven or insufficient clamping force. It offers fast response but is limited by clamping safety constraints and should not be adjusted drastically. Press-release cylinder stroke adjustment is suitable for press-release mechanisms where each circumferential position of the press-release cylinder can be independently controlled. It directly changes the press-release amount distribution at each circumferential position, offering high control precision, but is limited by the upper limit of a single press-release amount. In engineering practice, cylinder stroke difference adjustment is usually the primary method, supplemented by fine-tuning of clamping force. Both can be used individually or in combination, depending on the configuration of the on-site actuator. When a tilting jamming event is detected, the stroke is reduced at the circumferential position corresponding to the tilting direction and increased at the opposite circumferential position, creating circumferential asymmetric press-release compensation.

[0068] The various modules in the aforementioned intelligent smelting system for an electric arc furnace can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0069] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0070] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. An electrode pressing and discharging measuring device, characterized in that, The device includes: At least three non-contact displacement sensors are circumferentially distributed around the same electrode shell. Each of the non-contact displacement sensors is fixedly installed at different circumferential positions of the electrode holder to synchronously acquire the displacement signal of the electrode shell relative to the electrode holder at the corresponding circumferential position. The data processing unit receives the displacement signals collected by each of the non-contact displacement sensors and performs the following processing: The three-dimensional spatial attitude change of the electrode during the pressing and releasing process is calculated based on the displacement signals at at least three circumferential positions. The three-dimensional spatial attitude change includes at least the tilt direction and tilt angle of the electrode axis. Extract the variation characteristics of displacement signals at each circumferential position over time. If the displacement signals at all circumferential positions are detected to be monotonically increasing within the preset compression time window, and the total displacement at each circumferential position falls within the preset compression range, it is determined to be a real compression event and the compression amount is recorded. When a step back is detected in the displacement signal of at least one circumferential position outside the preset compression time window, or when the total displacement of the displacement signal of at least one circumferential position within the preset compression time window exceeds the upper limit of the preset compression range, the abnormal event that has occurred is determined to be an overall slippage event or a tilting and jamming event based on the three-dimensional spatial attitude change.

2. The apparatus according to claim 1, characterized in that, The calculation of the three-dimensional spatial attitude change of the electrode during the pressing and releasing process based on the displacement signals of the at least three circumferential positions includes: The mounting angle of each non-contact displacement sensor in the circumferential direction of the electrode shell and the corresponding real-time displacement value are obtained. Using the electrode holder as a reference, the spatial discrete point coordinates of the bottom end face of the electrode shell at each sampling time are constructed based on the installation angle and real-time displacement value of each sensor. Plane fitting is performed on the coordinates of each of the spatial discrete points to obtain the spatial plane equation of the bottom end face of the electrode shell, and the normal vector of the spatial plane is calculated. The normal vector is compared with the gravity direction vector, and the angle at which the normal vector deviates from the gravity direction is determined as the tilt angle of the electrode axis. The projection direction of the normal vector onto the horizontal plane is taken as the tilt direction.

3. The apparatus according to claim 1, characterized in that, The extraction of the time-varying characteristics of the circumferential displacement signals includes: Real-time sliding median filtering or mean filtering is performed on the displacement signals of each circumferential position to remove high-frequency noise interference; Perform a first-order difference operation on the filtered displacement signal to obtain the instantaneous displacement change rate at each sampling time. According to the preset time interval or number of sampling points, the filtered displacement signals of each circumferential position are accumulated and summed to obtain the total displacement of each circumferential position within any specified time period. The change characteristics include a real-time displacement value sequence, an instantaneous displacement change rate sequence, and the total displacement within a specified time period.

4. The apparatus according to claim 1, characterized in that, When the displacement signals at all circumferential positions are detected to be monotonically increasing within a preset release time window, and the total displacement at each circumferential position falls within a preset release range, it is determined to be a real release event and the release amount for that event is recorded, including: The moment when the instantaneous displacement change rate of at least one circumferential position is first detected to exceed a preset start threshold is taken as the start time of the compression and release time window, and a preset fixed duration is taken as the length of the compression and release time window. Within the compression and release time window, it is continuously determined whether the difference between the displacement signals of each circumferential position at adjacent sampling times is greater than or equal to zero. If a negative difference value appears in any channel, the determination is terminated. At the end of the pressing and releasing time window, the displacement difference between the end and start times of each circumferential position is calculated and used as the total displacement of that circumferential position in that cycle. The total displacement at each circumferential position in this cycle is compared with the lower and upper limits of the preset compression range, respectively. If the displacement signals of all circumferential positions within the window are monotonically increasing, and the total displacement of each circumferential position in this instance falls within the preset pressure release range, then it is determined to be a real pressure release event, and the average or weighted average of the total displacement of all circumferential positions in this instance is recorded as the pressure release amount for this instance.

5. The apparatus according to claim 1, characterized in that, Detecting a step back in the displacement signal at at least one circumferential position outside the preset pressing and releasing time window includes: Real-time monitoring of displacement signals in each circumferential position; when it is detected that the current moment is not within any triggered release time window, calculate the displacement difference between the current moment and the previous moment for each circumferential position. If the current displacement difference at a certain circumferential position is negative and its absolute value exceeds the preset step back threshold, and the displacement signal continues to decrease over multiple consecutive sampling times and the total cumulative decrease exceeds the preset cumulative back threshold, then it is determined that a step back has occurred at the circumferential position, and the event and the circumferential position where it occurred are recorded.

6. The apparatus according to claim 1, characterized in that, Detecting that the total displacement of the displacement signal at at least one circumferential position within the preset pressure release time window exceeds the upper limit of the preset pressure release range includes: Within each compression and release time window, the displacement of each circumferential position is accumulated in real time, and the real-time accumulated displacement is compared with the upper limit of the preset compression and release range; When the real-time cumulative displacement at any of the aforementioned circumferential positions exceeds the upper limit of the preset pressure release range, an over-range flag is immediately triggered, and the time corresponding to the triggering of the over-range flag and the information of the corresponding circumferential position are recorded. Monitoring continues until the end of the window.

7. The apparatus according to claim 1, characterized in that, The step of determining whether the currently occurring abnormal event is an overall slippage event or a tilting and jamming event based on the three-dimensional spatial posture change includes: When a step back event or a total displacement exceeding the limit event is detected, the tilt angle and tilt direction of the electrode axis at the current moment are obtained; The tilt angle is compared with a preset tilt threshold. If the tilt angle is less than the tilt threshold, and the displacement signals at each circumferential position change in the same direction and with similar amplitudes, then the current abnormal event is determined to be an overall slippage event. If the tilt angle is greater than or equal to the tilt threshold, and there are at least two circumferential positions where the displacement signals change in opposite directions or the displacement difference exceeds a preset inconsistency threshold, then the current abnormal event is determined to be a tilt jamming event.

8. A smart smelting system for an electric arc furnace, characterized in that, The system includes: The electrode pressing and discharging measuring device as described in claim 1; The smelting electrical parameter acquisition unit is used to acquire electrode current, furnace power and electrode-to-ground voltage in real time. The smelting data processing unit receives the discharge amount data, three-dimensional spatial attitude change and abnormal event judgment results output by the electrode discharge measurement device, and receives the electrical parameter data output by the smelting electrical parameter acquisition unit. The smelting data processing unit establishes a correlation model between electrode consumption rate and electrical parameters based on the pressure release sequence and electrical parameter sequence within a preset historical time window. Using the correlation model, it predicts the expected consumption rate of the electrode in real time based on the current electrical parameters and generates pressure release adjustment commands. The pressure-release equalization control unit receives the three-dimensional spatial attitude change and the abnormal event determination result. When the abnormal event determination result is a tilting and jamming event, it applies circumferentially differentiated pressure-release compensation to subsequent pressure-release operations by adjusting the clamping force of the gripper corresponding to each circumferential position or the stroke difference of each pressure-release cylinder, so that the electrode axis is restored to the preset verticality range. When the abnormal event determination result is an overall slippage event, it outputs a slippage alarm signal and triggers the reset of the pressure-release adjustment command.

9. The system according to claim 8, characterized in that, The process involves establishing a correlation model between electrode consumption rate and electrical parameters based on the pressure release sequence and electrical parameter sequence within a preset historical time window. This correlation model is then used to predict the expected electrode consumption rate in real time based on the current electrical parameters, and to generate pressure release adjustment commands, including: Obtain the discharge amount data and timestamps of each real discharge event recorded by the electrode discharge measurement device within a preset historical time window, and form a discharge amount time series; Acquire the electrode current sequence, furnace power sequence, and electrode-to-ground voltage sequence collected by the smelting electrical parameter acquisition unit within the preset historical time window, and align the electrical parameter sequence with the pressure release time sequence according to the timestamp. The actual electrode consumption rate in each time period is calculated based on the pressure release time series. The actual electrode consumption rate is used as the output label, and the average electrode current, average furnace power, and average electrode-to-ground voltage in the corresponding time period are used as input features to construct a training sample set. The preset regression model is trained using the training sample set to obtain the correlation model between the electrode consumption rate and electrical parameters. The regression model includes any one of multiple linear regression model, support vector regression model, or neural network regression model. The system acquires the real-time electrode current, real-time furnace power, and real-time electrode-to-ground voltage at the current moment. It then inputs these data into the trained correlation model and outputs the expected consumption rate of the electrode. Based on the expected consumption rate and the preset pressure release cycle, the expected pressure release amount required for the next pressure release cycle is calculated, and a corresponding pressure release adjustment command is generated.

10. The system according to claim 8, characterized in that, The step of applying circumferentially differentiated pressure-release compensation to subsequent pressure-release operations by adjusting the clamping force of the holder corresponding to each circumferential position or the stroke difference of each pressure-release cylinder, so as to restore the electrode axis to the preset perpendicularity range, includes: Receive the current three-dimensional spatial attitude change output by the electrode pressing and releasing measurement device, and obtain the current tilt direction and current tilt angle of the electrode axis; Determine whether the current tilt angle exceeds the preset verticality range. If it does, determine the circumferential position corresponding to the tilt direction based on the tilt direction. At the circumferential position corresponding to the tilting orientation, increase the preset increment value of the clamping force of the corresponding gripper, or decrease the preset reduction value of the stroke of the corresponding pressure-release cylinder to form a circumferentially asymmetrical clamping force distribution or pressure-release stroke distribution. At the relative circumferential position of the tilted orientation, reduce the preset reduction value of the clamping force of the corresponding gripper, or increase the preset increment value of the stroke of the corresponding pressure-release cylinder. The circumferentially differentiated clamping force adjustment amount and the pressure release cylinder stroke adjustment amount are written into the execution command of the next pressure release operation so as to apply circumferentially differentiated pressure release compensation during the next pressure release. After the next pressing and releasing operation is completed, the three-dimensional spatial attitude change output by the electrode pressing and releasing measurement device is acquired again, and it is determined whether the tilt angle of the electrode axis has recovered to the preset verticality range. If it has not recovered, the circumferential differential pressing and releasing compensation is iteratively executed until the electrode axis is recovered to the preset verticality range.

Images