Proportional valve linearization calculation system for electric furnace hydraulic lifting

Abstract

The present application relates to the field of industrial automation control and hydraulic servo control, in particular to a proportional valve linearization calculation system for electric furnace hydraulic lifting, comprising: synchronous acquisition of proportional valve driving voltage, electrode displacement, speed, hydraulic station pressure, furnace type identification, process state and electrode weight; reference voltage node matching and steady state judgment are performed at a 10 ms cycle, and a sampling trigger is output when the voltage falls within the tolerance and the pressure change rate and the speed change rate meet the limit value for 50 ms; signal loss, communication timeout or jump data are cleared and re-judged.

Description

Technical Field

[0001] This invention relates to the field of industrial automation control and hydraulic servo control, specifically to a linearization calculation system for a proportional valve used in the hydraulic lifting of an electric furnace. Background Technology

[0002] Existing electric furnace electrode hydraulic lifting control systems typically use proportional valves to drive hydraulic actuators to achieve continuous adjustment of electrode position; In related technologies, in order to improve the control accuracy of proportional valves, it is usually necessary to collect valve characteristic data under shutdown or special test conditions, and establish control parameters or correction tables accordingly; there are also solutions that directly use real-time sampling data to update parameters during operation to adapt to characteristic changes caused by valve core wear, oil aging, etc. However, shutdown calibration interrupts system operation, making it impossible to implement online dynamic closed-loop regulation. Directly using dynamic data can easily introduce invalid information such as transient states, communication anomalies, and process fluctuations, making it difficult to obtain stable characteristic parameters. At the same time, the direct overwriting of old parameters by new parameters during operation may cause abrupt changes in control output. Therefore, the existing methods for obtaining and updating proportional valve characteristic parameters cannot simultaneously ensure data reliability, online adaptability, and control stability. Summary of the Invention

[0003] The purpose of this invention is to provide a linearization calculation system for proportional valves used in hydraulic lifting of electric furnaces, thereby solving the technical problems in the prior art where directly using dynamic data easily introduces invalid information, making it difficult to stably obtain characteristic parameters, and where directly overriding old parameters easily causes sudden changes in control output. Therefore, the technical solution of this invention includes: The linearization calculation system for the proportional valve used in the hydraulic lifting of an electric furnace includes a main controller and a main control parameter storage area, a data acquisition module, a steady-state judgment module, a data cache module, a parameter calculation module, a parameter verification module, a smooth switching module, and a main control execution module. The data acquisition module obtains the proportional valve drive voltage, measured electrode speed, hydraulic station pressure, electrode weight, furnace type identification, and process status. The main controller presets discrete voltage nodes within the driving voltage range of the proportional valve as reference voltage nodes. The steady-state judgment module outputs a sampling trigger signal when the proportional valve drive voltage is within the preset tolerance of the reference voltage node, and the rate of change of the hydraulic station pressure and the measured electrode speed are both no greater than the corresponding limit and continue for a preset time. The data caching module stores the acquired data as sampled data according to the corresponding reference voltage node based on the sampling trigger signal. When the coverage of the reference voltage nodes with the sampled data reaches a preset lower limit, the parameter calculation module establishes a fitting model to generate new parameters containing the control voltage correction amount corresponding to each reference voltage node. When the parameter verification module determines that an update is allowed, it marks the new parameter as a parameter to be updated. When the safety conditions are met, the smooth switching module writes the parameter to be updated into the main control parameter storage area. The main control execution module calls the parameters in the main control parameter storage area to generate a proportional valve control signal, and outputs it to the proportional valve drive end after superimposing the real-time calculated process correction amount.

[0004] Preferably, the data acquisition module acquires electrode displacement and measured electrode velocity from an external displacement detection device, acquires hydraulic station pressure from an external hydraulic station pressure detection device, and acquires electrode weight, furnace type identification, and process status from an external host computer, manual input terminal, or process database; the data acquisition module includes a proportional valve drive voltage acquisition unit, an electrode displacement acquisition unit, an electrode velocity calculation unit, a hydraulic station pressure acquisition unit, a process status acquisition unit, a furnace type identification input unit, and an electrode weight input unit; The electrode displacement acquisition unit receives the displacement signal output by the displacement detection device, the hydraulic station pressure acquisition unit receives the pressure signal output by the hydraulic station pressure detection device, and the furnace type identification input unit, process status acquisition unit, and electrode weight input unit receive corresponding data provided by the host computer, manual input terminal, or process database.

[0005] Preferably, the steady-state judgment module calculates the hydraulic station pressure change rate and the measured electrode speed change rate according to a fixed sampling period; when the proportional valve drive voltage is within the preset tolerance of the reference voltage node, the hydraulic station pressure change rate is not greater than the preset pressure change rate limit, the measured electrode speed change rate is not greater than the preset speed change rate limit, and the above conditions are continuously satisfied for a preset stable duration, the corresponding reference voltage node sampling trigger signal is output.

[0006] Preferably, the reference voltage node includes a range of -10V to +10V. The system includes 41 equally spaced nodes with a step size and including the endpoints. In addition to these 41 equally spaced nodes, a proportional valve dead zone node, obtained through a proportional valve no-response zone test, is added. The total number of reference voltage nodes is denoted as […]. ; The proportional valve dead zone node includes at least one of a forward opening voltage node, a reverse opening voltage node, and a zero-position node; the preset tolerance is... to The preset pressure change rate limit is to , where the preset speed change rate limit value is to ; where the preset stable duration is .

[0007] Preferably, the data cache module is provided with an independent storage area for each reference voltage node, and the independent storage area stores a node voltage identifier, a node average electrode speed, a node average electrode weight, a node average hydraulic station pressure, a furnace type identifier statistical value and an effective sample count; when sampling data is obtained for the corresponding reference voltage node, the exponential moving average method is adopted to update the node average electrode speed, the node average electrode weight and the node average hydraulic station pressure, and the furnace type identifier statistical value is updated according to the furnace type identifier with the highest occurrence frequency, where the forgetting factor of the exponential moving average is to .

[0008] Preferably, the data cache module is further configured to reject abnormal data; when the difference between the measured electrode speed in the sampling data and the node average electrode speed of the corresponding reference voltage node exceeds the preset speed tolerance of the reference voltage node, the sampling data is not written into the corresponding independent storage area; the preset speed tolerance is determined according to the design target electrode speed and the speed detection accuracy, and the design target electrode speed is determined by the design motion curve pre-stored in the main control parameter storage area.

[0009] Preferably, the parameter calculation module calculates the effective sampling node coverage, where the effective sampling node coverage is the ratio of the number of nodes whose effective sample count reaches the minimum number of effective samples per node to the total number of reference voltage nodes; the minimum number of effective samples per node is 5 to 50, and the preset coverage lower limit is 0.6 to 0.9; When the effective sampling node coverage reaches the preset coverage lower limit, a multiple linear fitting model is established based on the driving voltage, electrode weight, hydraulic station pressure, corresponding coded value of the furnace type identifier and measured electrode speed corresponding to the effective nodes, where the corresponding coded value of the furnace type identifier is determined by the fixed coding table pre-stored in the main control parameter storage area.

[0010] Preferably, the parameter calculation module divides the voltage range from -10V to +10V into a high-speed descending zone, a low-speed descending zone, a dead zone and inching zone, a low-speed ascending zone and a high-speed ascending zone according to preset voltage thresholds U1, U2, U3 and U4, where U1 is the boundary threshold between the high-speed and low-speed descending zones, U2 is the boundary threshold between the low-speed descending zone and the dead zone, U3 is the boundary threshold between the dead zone and the low-speed ascending zone, U4 is the boundary threshold between the low-speed and high-speed ascending zones, and U1<U2<U3<U4, and the threshold boundaries of each zone are determined according to the preset values pre-written in the main control parameter storage area; The maximum rise rate and maximum fall rate are process parameters pre-stored in the main control parameter storage area, and the designed motion curve is a piecewise linear curve established based on the maximum rise rate, maximum fall rate and drive voltage range.

[0011] Preferably, the parameter calculation module generates the control voltage correction amount for each reference voltage node based on the deviation between the design target electrode velocity given by the design motion curve and the model predicted electrode velocity output by the fitted model. The design target electrode velocity is determined by the design motion curve pre-stored in the main control parameter storage area. When the design target electrode speed difference between adjacent nodes is not zero, the control voltage correction amount is generated using the correction gain determined by the slope of the adjacent nodes. When the design target electrode speed difference between adjacent nodes is zero, the control voltage correction amount is generated using the preset correction gain. The boundary nodes use the preset boundary correction gain. The main control execution module performs linear interpolation of the control voltage correction amount between adjacent reference voltage nodes based on the current driving voltage.

[0012] Preferably, the parameter verification module determines the parameter qualification based on the weighted evaluation results of effective sampling node coverage, sample velocity standard deviation, and root mean square error. When the parameter qualification reaches a preset qualification threshold, the new fitting coefficient and node lookup correction amount output by the parameter calculation module are marked as parameters to be updated. The preset qualification threshold is 0.7 to 0.95. The smooth switching module starts writing when any of the following preset safe switching conditions are met: the process state is the tapping period or the feeding period and the electrode is in the highest locked state; or the measured absolute value of the electrode speed is not greater than the preset safe speed threshold and continues to reach the preset safe duration; the smooth switching module interpolates and switches the old parameter vector and the parameter vector to be updated according to the preset transition coefficient, and generates a proportional valve control output during the switching period; The process correction amount is determined based on the deviation between the current electrode weight and the historical average electrode weight, the deviation between the current hydraulic station pressure and the historical average hydraulic station pressure, and the difference between the corresponding code value of the furnace type identifier and the reference code value pre-stored in the main control parameter storage area. The relevant correction coefficients are stored in the main control parameter storage area, and the final output is limited to the allowable drive voltage range of the proportional valve.

[0013] Compared with the prior art, the present invention has the following beneficial effects: 1. By synchronously collecting drive voltage, displacement, pressure, and process input data during normal electric furnace production using only local stable conditions, and combining this with steady-state screening based on voltage tolerance, pressure change rate, velocity change rate, and continuous stable duration, we can avoid dedicated calibration during furnace shutdown and reduce interference from transient state, communication anomalies, and mechanical inertia data on characteristic modeling, thereby improving the reliability and effectiveness of proportional valve steady-state sample acquisition. Furthermore, by handling packet loss, communication interruption, time stamp anomalies, and signal failures through abort and reset mechanisms, we can prevent erroneous samples from triggering sampling and improve the continuity and stability of the data link. 2. By writing steady-state samples into independent storage areas according to the reference voltage nodes, and updating the node representative values ​​using exponential moving average for the valid samples after anomaly removal, while retaining the original valid velocity sample statistical information for standard deviation calculation, the consistency of sample reliability evaluation can be maintained while suppressing slow drift caused by mechanical wear and oil aging. By reconstructing the partition characteristic model only after the node coverage, sample dispersion, and fitting error all meet the conditions, and retaining the parameters of the previous period when the conditions are not met or the coverage is insufficient, the distortion of the global model caused by local data loss can be avoided. 3. By gradually switching between old and updated parameters using linear interpolation within the safety process window after parameter verification, and by performing output limiting and status interruption recovery processing during the switching process, the sudden changes and oscillations in the control benchmark caused by the new parameters directly overwriting the old parameters can be avoided, ensuring continuous and stable proportional valve control output. By writing the updated parameters into the main control parameter storage area only after the pass rate reaches the preset threshold, the safety and stability of online updates can be further improved. Attached Figure Description

[0014] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. The drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0015] Figure 1 This is a schematic diagram of the functional modules of the linearization calculation system for the proportional valve of the electric furnace hydraulic lifting system provided in this application embodiment. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0017] Please see the appendix Figure 1The linearization calculation system for the proportional valve of the electric furnace hydraulic lifting includes a main controller and a main control parameter storage area, a data acquisition module, a steady-state judgment module, a data cache module, a parameter calculation module, a parameter verification module, a smooth switching module, and a main control execution module. The data acquisition module obtains the proportional valve drive voltage, measured electrode speed, hydraulic station pressure, electrode weight, furnace type markings, and process status. The main controller presets discrete voltage nodes within the proportional valve drive voltage range as reference voltage nodes. The steady-state judgment module outputs a sampling trigger signal when the proportional valve drive voltage is within the preset tolerance of the reference voltage node, and the rate of change of the hydraulic station pressure and the measured electrode speed are both no greater than the corresponding limit and continue for a preset time. The data caching module stores the acquired data as sampled data according to the corresponding reference voltage node based on the sampling trigger signal; When the coverage of the reference voltage nodes with sampled data reaches a preset lower limit, the parameter calculation module establishes a fitting model to construct new parameters containing the control voltage correction amount corresponding to each reference voltage node. When the parameter validation module determines that an update is allowed, it marks the new parameter as a parameter to be updated. The smooth switching module writes the parameters to be updated into the main control parameter storage area when safety conditions are met; The main control execution module calls the parameters in the main control parameter storage area to construct the proportional valve control signal, and outputs it to the proportional valve drive end after superimposing the real-time calculated process correction amount.

[0018] Based on the above embodiments, further: the data acquisition module obtains electrode displacement and measured electrode velocity from an external displacement detection device, obtains hydraulic station pressure from an external hydraulic station pressure detection device, and obtains electrode weight, furnace type identification, and process status from an external host computer, manual input terminal, or process database; the data acquisition module includes a proportional valve drive voltage acquisition unit, an electrode displacement acquisition unit, an electrode velocity calculation unit, a hydraulic station pressure acquisition unit, a process status acquisition unit, a furnace type identification input unit, and an electrode weight input unit; The electrode displacement acquisition unit receives the displacement signal output by the displacement detection device, the hydraulic station pressure acquisition unit receives the pressure signal output by the hydraulic station pressure detection device, and the furnace type identification input unit, process status acquisition unit, and electrode weight input unit receive the corresponding data provided by the host computer, manual input terminal, or process database. The steady-state judgment module calculates the hydraulic station pressure change rate and the measured electrode speed change rate according to a fixed sampling period. When the proportional valve drive voltage is within the preset tolerance of the reference voltage node, the hydraulic station pressure change rate is not greater than the preset pressure change rate limit, the measured electrode speed change rate is not greater than the preset speed change rate limit, and the above conditions are continuously satisfied for a preset stable duration, the corresponding reference voltage node sampling trigger signal is output. The reference voltage nodes are included in the range of -10V to +10V. The system includes 41 equally spaced nodes with a step size set and including the endpoints. In addition to these 41 equally spaced nodes, a proportional valve dead zone node, obtained through a proportional valve no-response zone test, is added. The total number of reference voltage nodes is denoted as […]. ; The proportional valve dead zone node includes at least one of a forward opening voltage node, a reverse opening voltage node, and a zero-position node; the preset tolerance is... to The preset pressure change rate limit is to The preset speed change limit is to The preset stable duration is ; This implementation method is mainly used to illustrate the basic hardware architecture and steady-state data acquisition logic of the linearization calculation system for the proportional valve of the electric furnace hydraulic lifting. Its basic design idea is to avoid the need for special parameter calibration during furnace shutdown, and instead use the brief local stable conditions during normal electric furnace production to extract effective proportional valve characteristic data. For example, in actual application conditions, take the electrode hydraulic lifting control system of a refining furnace or electric arc furnace as an example; during continuous smelting, the electrode needs to be frequently raised and lowered to adjust the arc length, and the hydraulic system is in a high-frequency dynamic response state; the main controller is equipped with a main control parameter storage area to store the currently effective control parameters. The data acquisition module serves as the system's data access point, synchronizing various physical quantities on-site in real time. The proportional valve drive voltage acquisition unit acquires the voltage signal output by the control loop, the electrode displacement acquisition unit reads the signal from the laser rangefinder or displacement sensor, the electrode speed calculation unit performs differential calculation on the displacement signal to obtain the measured electrode speed, and the hydraulic station pressure acquisition unit acquires the pipeline pressure. Simultaneously, the furnace type identification input unit, process status acquisition unit, and electrode weight input unit read the furnace type identification, process status, and electrode weight of the current furnace batch from the factory process database, the host computer, or a manual input terminal, respectively. The above multi-dimensional data are aggregated to the main controller according to the control scanning cycle. After acquiring the continuous data stream, directly using all dynamic data for characteristic modeling would introduce a large amount of invalid data due to hydraulic transients and mechanical inertia. Therefore, the data stream is filtered by the steady-state judgment module. This module pre-divides discrete reference voltage nodes, including... to Within the range 41 equally spaced nodes with step size distribution, and dead zone nodes such as forward opening, reverse opening, and zero position calibrated through previous tests; When multiple reference voltage nodes simultaneously fall within the preset voltage tolerance range, the steady-state judgment module determines the corresponding node based on the principle of minimizing the absolute deviation between the driving voltage and the node voltage. When the absolute deviations are the same, dead-zone nodes are selected first, thus ensuring that each valid sample is written to only one independent storage area. The steady-state judgment module follows a fixed... The sampling period is used to calculate the current rate of change of hydraulic station pressure and the measured rate of change of electrode speed. When the drive voltage of a certain cycle falls into a certain reference voltage node setting to When the preset voltage tolerance is within the specified range, the module initiates a steady-state condition check; if the calculated rate of change of hydraulic station pressure at this time is not greater than... to The limit, and the measured rate of change of electrode velocity is not greater than to The limit indicates that the hydraulic station pressure and electrode speed are in a local stable state within the current sampling window, and can be used as steady-state characterization data of the driving voltage node, rather than directly determining that the hydraulic valve core has reached absolute stillness or complete equilibrium. To eliminate transient interference, the module requires that the above conditions be met continuously. That is, five consecutive control scan cycles; once the preset stable duration is reached, the steady-state judgment module outputs a sampling trigger signal to the data buffer module, marking the data at the current moment as a valid steady-state sample; To ensure the continuity of the data link and the stability of the system, if the data acquisition module loses sensor signals or the fieldbus communication times out within a continuous judgment period, the steady-state judgment module will directly stop the steady-state cumulative count of the current node, clear the duration state, and restart the judgment in the next valid data period, thereby avoiding the erroneous triggering of the sampling signal due to data loss or communication interruption. During periods of data anomaly, the process status acquisition unit synchronously marks the current process status as invalid. Only after the proportional valve drive voltage, displacement signal, hydraulic station pressure, and process input data have all been restored to validity will the reference voltage node matching and steady-state duration accumulation be performed again. Among them, the preset voltage tolerance is primarily determined based on the proportional valve drive output resolution, drive voltage acquisition accuracy, and proportional valve dead zone test results, and is also... to The range is selected; the pressure change rate limit is determined based on the measurement noise of the hydraulic station pressure sensor, the allowable fluctuation of pressure regulation, and the pressure stability required for the steady-state sample, and is within the specified range. to Select from the range; The limit for the rate of change of velocity is determined based on the resolution of the displacement detection device, the velocity calculation error, and the allowable acceleration fluctuations of the electrode lifting control, and is... to The above limits are selected within the specified range; they are written into the main control parameter storage area during system initialization, and are called by the steady-state judgment module during operation, and remain unchanged for the same furnace type and the same control mode. In practice, the data acquisition module and the steady-state judgment module execute according to a fixed control scan task, each... The cycle first completes the sampling and latching of each input quantity at the same moment, and then the electrode velocity calculation unit calculates the velocity based on two adjacent effective displacement sampling values, and sends the data of this cycle with a unified sampling time mark to the steady state judgment module. The data acquisition module employs an alternating data buffering method: the current cycle only writes to the currently used sampling buffer, while the buffer sealed in the previous cycle is available for the steady-state judgment module to read; at the cycle boundary, the handover is completed by switching the buffer pointer once, and the steady-state judgment module must not read data that has not yet been sealed; if the calculation task of a certain cycle is not completed before the start of the next sampling cycle, no new steady-state trigger signal is generated in that cycle, and unprocessed data is not queued indefinitely, but is overwritten by the complete sampling result of the next cycle, thus ensuring that the execution time of the control task is controlled. Periodic constraints; The steady-state judgment module operates in the following state sequence: invalid, pending confirmation, continuous stability, triggered, and reconfirmation. When the drive voltage node is successfully matched and all inputs are valid, it changes from the invalid state to the pending confirmation state. When both the pressure change rate and the velocity change rate meet the limits, the continuous stability count is incremented by one; otherwise, it is immediately cleared to zero and returns to the pending confirmation state. After five consecutive counts, the module transitions to a triggered state, outputting only one sampling trigger signal and maintaining the triggered state until the drive voltage leaves the node tolerance range or any input fails, at which point it transitions to a reconfirmation state. When the fieldbus times out, the time stamp goes backward, or the displacement signal jumps beyond the physical travel, it is treated as an input failure, and the continuous count of the current node is cleared. For repeated sampling results arriving within the same period, the module only retains the more recent complete result to avoid the same physical moment being counted repeatedly.

[0019] The data cache module sets an independent storage area for each reference voltage node, and the independent storage area stores the node voltage identifier, the node average electrode speed, the node average electrode weight, the node average hydraulic station pressure, the furnace type identifier statistical value and the valid sample count; When the corresponding reference voltage node obtains sampling data, the exponential moving average method is adopted to update the node average electrode speed, the node average electrode weight and the node average hydraulic station pressure, and the furnace type identifier statistical value is updated according to the furnace type identifier with the highest occurrence frequency, the forgetting factor of the exponential moving average is to ; The data cache module is further configured to reject abnormal data; when the difference between the measured electrode speed in the sampling data and the node average electrode speed of the corresponding reference voltage node exceeds the preset speed tolerance of the reference voltage node, the sampling data is not written into the corresponding independent storage area; the preset speed tolerance is determined according to the design target electrode speed and the speed detection accuracy, and the design target electrode speed is determined by the design motion curve pre-stored in the main control parameter storage area; The parameter calculation module calculates the valid sampling node coverage, which is the ratio of the number of nodes whose valid sample count reaches the minimum number of valid samples per node to the total number of reference voltage nodes; the minimum number of valid samples per node is 5 to 50, and the preset lower coverage limit is 0.6 to 0.9; When the valid sampling node coverage reaches the preset lower coverage limit, a multiple linear fitting model is established based on the driving voltage, electrode weight, hydraulic station pressure, the coded value corresponding to the furnace type identifier and the measured electrode speed corresponding to the valid nodes, and the coded value corresponding to the furnace type identifier is determined by the fixed coding table pre-stored in the main control parameter storage area; The parameter calculation module divides the voltage range from -10V to +10V into a high-speed descending zone, a low-speed descending zone, a dead zone and micro-motion zone, a low-speed ascending zone and a high-speed ascending zone according to preset voltage thresholds U1, U2, U3 and U4, where U1 is the boundary threshold between the high-speed and low-speed descending zones, U2 is the boundary threshold between the low-speed descending zone and the dead zone, U3 is the boundary threshold between the dead zone and the low-speed ascending zone, U4 is the boundary threshold between the low-speed and high-speed ascending zones, and U1<U2<U3<U4, the threshold boundary of each zone is determined according to the preset value pre-written in the main control parameter storage area; The maximum ascending speed and the maximum descending speed are process parameters pre-stored in the main control parameter storage area, and the design motion curve is a piecewise linear curve established according to the maximum ascending speed, the maximum descending speed and the driving voltage range; The parameter calculation module constructs the control voltage correction amount of each reference voltage node according to the deviation between the design target electrode speed given by the design motion curve and the model predicted electrode speed output by the fitting model, and the design target electrode speed is determined by the design motion curve pre-stored in the main control parameter storage area; When the design target electrode speed difference between adjacent nodes is not zero, the control voltage correction amount is constructed using the correction gain determined by the slope of the adjacent nodes. When the design target electrode speed difference between adjacent nodes is zero, the control voltage correction amount is constructed using the preset correction gain. The boundary nodes use the preset boundary correction gain. The main control execution module performs linear interpolation of the control voltage correction amount between adjacent reference voltage nodes based on the current driving voltage. This implementation mainly explains how the system cleans and accumulates the discretely acquired steady-state data, and reconstructs the characteristic model of the proportional valve when the data volume meets the requirements; if the slow drift caused by valve core wear and oil aging during long-term operation is not modeled and corrected, the measured output flow of the proportional valve will gradually deviate from the design expectation. Based on the operating conditions of the previous implementation, when the data cache module receives the sampling trigger signal from the steady-state judgment module, the system does not directly use the single sample, but instead diverts it to the independent storage area of ​​the corresponding reference voltage node; before data is written, the data cache module performs anomaly removal and interception. For nodes without valid samples, the first data obtained through steady-state judgment is directly used to initialize the average electrode speed, average electrode weight, and average hydraulic station pressure of the node, and the valid sample count is initialized to 1; for nodes with valid samples, the system reads the design motion curve in the main control parameter storage area, obtains the design target electrode speed of the node, and determines the preset speed tolerance in combination with the speed detection accuracy. If the deviation between the measured electrode velocity in the current sampled data and the average electrode velocity of existing nodes in the independent storage area exceeds the tolerance, it indicates that the sample may be affected by mechanical collision or electromagnetic interference, and the system directly removes the sample. For valid samples that pass the verification, the data caching module updates the average electrode velocity, average electrode weight, and average hydraulic station pressure of the node using the exponential moving average (EMA) algorithm. to The forgetting factor ensures that the system retains the stability of historical data while enabling the updated average value to smoothly follow the slow-changing characteristics caused by mechanical wear. Simultaneously, the valid sample count of the node is updated, and the furnace type identifier with the most occurrences is updated to the statistical value of the furnace type identifier of the node according to the number of times each furnace type identifier appears in the valid samples of the node; when the number of occurrences is the same, the furnace type identifier determined in the previous statistical period is retained, and when they are tied for the first time, they are determined according to the preset furnace type code order; the code value corresponding to the furnace type identifier is converted from the fixed code table pre-stored in the main control parameter storage area; To support the subsequent parameter verification module in calculating the standard deviation of sample velocity, each independent storage area, while storing the representative value of the storage node and the effective sample count, retains the statistical information of the original effective samples of the measured electrode velocities that have been rejected due to anomalies. The statistical information includes at least the cumulative sum of velocity samples, the cumulative sum of velocity squares, and the corresponding effective sample count, or it can be stored using an equivalent online variance accumulation. The statistical information is updated in the order in which samples are removed due to anomalies, and the sample dispersion is not calculated after EMA smoothing of the measured electrode speed; samples removed due to communication anomalies, sensor failure, or speed deviation exceeding the limit are not included in the statistical information. Therefore, the average electrode velocity of a node is used to characterize the slowly varying representative value of a node, and the statistical information of the original effective velocity samples is used to calculate the standard deviation of the sample velocity of that node. The two respectively undertake the functions of model fitting and parameter reliability evaluation. The standard deviation of the sample velocity is only calculated for nodes with an effective sample count of not less than 2. In this embodiment, all effective nodes that reach the minimum number of effective samples per node meet this condition, avoiding the standard deviation being undefined due to insufficient sample count. Specifically, for the current sample that has passed the anomaly removal, the node representative value is updated according to the weighted sum of the current sample and the previous node representative value. The update rule is as follows: in, This is the average value of the nodes before the update. The current valid sample value, This is the updated node average. As a forgetting factor; the electrode speed, electrode weight, and hydraulic station pressure are each updated independently once, and the first valid sample is used directly as the initial value without weighting. The sample count is increased only after the sample passes the speed deviation verification and the above node values ​​are updated. Samples that are removed due to communication abnormalities, sensor malfunctions, or speed deviation exceeding the limit are not included in the valid sample count. The speed sample statistics are updated synchronously with the sample count to ensure that the calculation range of the sample speed standard deviation is consistent with the valid sample count. As the electric furnace production cycle progresses, data from each independent storage area gradually accumulates; the parameter calculation module periodically scans the effective sample count of each node; when the count of a node reaches the lower limit requirement of 5 to 50, it is marked as a valid node; the parameter calculation module calculates the coverage of valid sampling nodes, where the number of valid nodes is based on the total number of reference voltage nodes. Perform statistics on the denominator; When the coverage reaches to When the preset lower limit is reached, it indicates that the currently accumulated data has the ability to represent the global characteristics of the proportional valve, triggering the model reconstruction logic. During model reconstruction, the node representative data in the independent storage area of ​​each effective node after anomaly removal and exponential moving average processing is used as the fitting sample. Each effective node corresponds to one node representative data to avoid mistaking the repeated sampling times of the same node as independent fitting samples. Before the formal fitting, the parameter calculation module also checks whether the driving voltage, electrode weight, hydraulic station pressure and furnace type identification code corresponding to the effective node have a preset minimum change range; when the change range of any input quantity is lower than the corresponding minimum change range, or the fitting input data formed by the effective node is insufficient to support the stable determination of the fitting parameters, the fitting is suspended and the model parameters of the previous cycle are retained. During the modeling process, the parameter calculation module is based on a preset voltage threshold. to The total voltage range is divided into a high-speed descent zone, a low-speed descent zone, a dead zone and micro-motion zone, a low-speed rise zone, and a high-speed rise zone. For each zone, the module extracts the driving voltage, electrode weight, hydraulic station pressure, and furnace type identification code value converted by a fixed coding table from the independent storage area of ​​the effective node as input, and takes the measured electrode speed as output. It then performs multiple linear fitting calculations to construct the fitting model parameters for each zone. Specifically, the parameter calculation module first establishes an input data table according to the interval to which the voltage node belongs. Each row corresponds to a valid node, and each column corresponds to the driving voltage, electrode weight, hydraulic station pressure and furnace type identification code value in turn. The last column is the average electrode speed of the node. The least squares method is used to obtain the fitting coefficients and constant terms corresponding to each input quantity, and the fitting residual is used to check whether the coefficients can be stably determined. Since the fitting model for each interval contains four inputs and one constant term, the effective node data for that interval should form at least five linearly independent design matrices. When the number of effective nodes in an interval is insufficient to form full-rank input data, or when the range of change of any input column is lower than the corresponding minimum range of change, the new fitting parameters for that interval are not output, but the fitting parameters for the corresponding interval of the previous period are retained. Dead zone nodes located outside the 41 equally spaced nodes still participate in node storage and coverage statistics, and are assigned to the fitting processing of dead zone and micro-motion zone according to the preset boundary assignment rules. For nodes that do not reach the minimum number of effective samples per node, they are not used as fitting samples for the corresponding interval. The module substitutes each reference voltage node into the newly established fitting model to obtain the model's predicted electrode velocity, and compares it with the design target electrode velocity in the design motion curve to calculate the deviation. Based on the deviation and the correction gain determined by the slope of adjacent nodes, the preset correction gain is called when the difference between the target velocities of adjacent nodes is zero, and the preset boundary correction gain is called at boundary nodes to construct the control voltage correction amount for each reference voltage node. Among them, the slope of adjacent nodes is determined according to the difference between the design target electrode speed and the difference between the driving voltage of adjacent reference voltage nodes, and the corresponding correction gain is called according to the descent, dead zone and micro-motion or rise interval to which it belongs. The correction gain is determined by the degree of response of the control voltage change to the electrode speed deviation in the proportional valve historical adjustment test, and is limited to the allowable gain range preset in the main control parameter storage area; when the design target electrode speed difference between adjacent nodes is zero, or when the node is located at the voltage interval boundary, the corresponding preset correction gain or preset boundary correction gain is called respectively to avoid division by zero and boundary abrupt changes in the slope calculation; Specifically, for the first For each reference voltage node, the speed deviation of the node is obtained by subtracting the model-predicted electrode speed from the target electrode speed. Then, the speed deviation is multiplied by the signed correction gain determined based on the interval, the slope of adjacent nodes, and historical adjustment tests to obtain the unlimited control voltage correction amount. The physical dimension of the signed correction gain is the ratio of control voltage to electrode speed. Its value is determined by converting the slope of the target speed of adjacent nodes and combining it with the results of historical adjustment tests, so as to ensure that the control voltage is obtained by multiplying the speed deviation and the correction gain. When the correction amount exceeds the single-node correction amount allowed range preset in the main control parameter storage area, it is truncated according to the upper or lower limit of the allowed range, and finally the control voltage correction amount written into the node lookup table data is obtained. The sign of the signed correction gain is consistent with the direction of the measured influence of the driving voltage change on the electrode speed within the corresponding interval, so that when the model predicts the speed is lower than the design target speed, the correction amount is output in the direction of increasing the electrode speed; the correction amount between adjacent nodes is linearly interpolated by the main control execution module according to the ratio of the current driving voltage to the voltage of the two nodes, and when the current driving voltage falls on the node endpoint, the correction amount of the corresponding endpoint is directly used. During data accumulation and modeling, if the electric furnace operates in a specific process section for a long time, such as being in a low-speed fine-tuning state for a long time, the number of effective samples of nodes in the high-speed section will not reach the minimum number of effective samples of a single node, making the overall coverage unable to meet the lower limit. The parameter calculation module will suspend the global fitting task and retain the model parameters of the previous cycle, thereby avoiding global model distortion caused by the lack of local data. In practice, the independent storage areas of each reference voltage node are managed logically as empty nodes, accumulated nodes, valid nodes, nodes to be rebuilt, and frozen nodes. The sampled data will only move the corresponding node from an empty node or accumulated node to the next state after node matching, integrity check, speed deviation verification, and node value update are completed. When communication is abnormal, the sensor is invalid, the count is insufficient, or the fitting interval does not meet the full rank condition, the corresponding node remains in its original state and the current data is not converted into a valid node; the data caching module first completes the abnormal removal of the sampling results within a control cycle, and then performs a node value update and a statistical update once for the same node; the parameter calculation module only reads the node data after the node update is completed. To prevent the main control execution module from reading intermediate results that are being updated, the data caching module adopts a node-level mutual exclusion access and version switching method: during the update, the average value, furnace type identification statistics, effective sample count and original speed statistics are first updated in the temporary copy to ensure consistency. After verifying that the count corresponds to the statistics, the temporary copy is switched to the current copy all at once. When scanning, the parameter calculation module reads all copies of the same version of nodes and does not allow overwriting of this version during the scan. If the scanning process detects that the node version is inconsistent, the effective sample count is inconsistent with the statistical count, or the fitting task has not been completed for more than one calculation cycle, the scan results are discarded and the model parameters of the previous cycle are retained, without blocking the control cycle of the main control execution module. For the effective sample count is , velocity sample cumulative sum is The cumulative sum of squared velocity is At the node, the parameter verification module calculates the standard deviation of the sample velocity based on the original valid velocity samples, specifically as follows: When a value less than zero appears within the square root due to the finite word length, and its absolute value is not greater than the calculation precision tolerance, it is treated as zero. If it exceeds the tolerance, the statistical information is judged to be abnormal and the node is frozen. The statistical information of each node is only updated after the data that has been removed through anomaly removal is written. The representative value of the EMA node does not participate in the above dispersion calculation, thereby ensuring that the data sources used for the fitted representative value, sample size and credibility evaluation are consistent with each other.

[0020] The parameter verification module determines the parameter passability based on a weighted evaluation of effective sampling node coverage, sample velocity standard deviation, and root mean square error. When the parameter passability reaches a preset passability threshold, the new fitting coefficients and node lookup corrections output by the parameter calculation module are marked as parameters to be updated. The preset passability threshold is 0.7 to 0.95. The smooth switching module initiates writing when any of the following preset safe switching conditions are met: The process is in the tapping or feeding phase and the electrode is in the highest locked position; or the measured absolute value of the electrode speed is not greater than the preset safe speed threshold and continues to reach the preset safe duration; the smooth switching module interpolates and switches the old parameter vector and the parameter vector to be updated according to the preset transition coefficient, and constructs a proportional valve control output during the switching period; The process correction amount is determined based on the deviation between the current electrode weight and the historical average electrode weight, the deviation between the current hydraulic station pressure and the historical average hydraulic station pressure, and the difference between the corresponding code value of the furnace type identifier and the reference code value pre-stored in the main control parameter storage area. The relevant correction coefficients are stored in the main control parameter storage area, and the final output is limited to the allowable drive voltage range of the proportional valve. This implementation mainly explains how to assess the reliability of the newly constructed proportional valve characteristic parameters and replace the main control parameters without disturbance without affecting the normal smelting of the electric furnace. If the parameters newly calculated by the parameter calculation module are directly overwritten in the storage area of ​​the main control parameters that are currently in operation, it will cause a jump in the control reference during the electrode adjustment process, resulting in the electrode acceleration falling below the preset limit or control oscillation. Following the previous implementation method, after the parameter calculation module constructs new fitting coefficients and node lookup correction amounts, the data is transferred to the parameter verification module. The verification module extracts the node coverage of this modeling, the standard deviation of the sample velocity calculated by each effective node based on the statistical information of the original effective velocity samples after anomaly removal, and the root mean square error of the fitting model, and converts them into dimensionless evaluation scores according to the preset qualified range. Each evaluation score is normalized according to the preset upper and lower limits of the corresponding indicator. Within the acceptable range, scores are continuously converted according to the direction of indicator improvement. Scores exceeding the acceptable range are limited to the upper or lower limit, respectively. to Between these parameters, the higher the node coverage, the smaller the sample velocity standard deviation, and the smaller the root mean square error, the higher the corresponding evaluation score. The standard deviation of the sample velocity is calculated based on the statistical information of the original effective velocity samples of each effective node. The average electrode velocity of the node after EMA update is not used repeatedly as a sample to avoid misclassifying the representative value of the same node as multiple independent samples. For the standard deviation of the sample velocity of each effective node, the parameter verification module performs weighted summation according to the effective sample count to form a single standard deviation index for evaluation. The root mean square error is calculated by summing the residuals between the representative data of each effective node in each interval and the corresponding fitted output, and is uniformly calculated according to the number of fitted effective nodes. Higher node coverage, smaller sample velocity standard deviation, and smaller root mean square error correspond to higher evaluation scores. The verification module then performs weighted calculations according to preset weights to determine the parameter pass rate. The sum of the preset weights for each parameter is 1 to avoid evaluation distortion caused by directly weighting different physical quantities due to differences in dimensions and numerical ranges. If the pass rate reaches... to If the preset threshold is reached, it indicates that the new model covers a sufficient working range and has good data convergence. The module then marks these parameters as parameters to be updated. If the preset threshold is not reached, the current master control parameters are retained and the subsequent write process is not started. Specifically, let the measured value of a certain indicator to be evaluated be... The lower and upper limits of its qualified range are respectively and For node coverage where a higher indicator value is better, the evaluation score is based on the indicator's performance within the specified range. to Calculation of relative position within the interval; for samples where smaller index values ​​are preferred, the standard deviation and root mean square error are calculated based on their relative position to the upper limit of the acceptable range. Results exceeding the acceptable range are truncated. or ; in, and The parameters are pre-determined based on historical stable operating data, sensor accuracy, and model allowable error, and are written into the main control parameter storage area. The three evaluation scores are weighted with fixed weights. The weights are determined during system calibration based on the influence of node coverage, data dispersion, and fitting error on the reliability of parameter updates, and remain unchanged under the same control mode. When any sub-item evaluation score cannot be calculated due to missing data, or when there are intervals in the fitted model that are not included in the evaluation, the parameter verification module directly determines that the parameters in this batch are unqualified and does not compensate with the higher scores of other sub-items. To ensure that the evaluation scores have defined calculation boundaries, the score corresponding to node coverage is calculated based on the relative position of its measured value between the lower and upper limits of the acceptable range, and the calculation results are restricted to... to Between; the scores corresponding to the sample velocity standard deviation and root mean square error are calculated based on the relative position of the measured value from the upper limit of the pass, and the calculation results are also limited to... to between; The parameter pass rate is obtained by multiplying the scores of the three sub-items by their corresponding fixed weights and then summing them; if any indicator's score is... Not greater than If the indicator data is missing, the node coverage cannot represent all the intervals to be updated, or no evaluable fitting results are constructed for the corresponding interval, then the score calculation for this item will not be performed and the parameters of this batch will be directly judged as unqualified. The smooth switching module enters a waiting state, continuously monitoring the electric furnace operating information from the process status acquisition unit; the module will not start updating during the melting period of high-frequency electrode adjustment, but waits for a safe process window; when the system detects that the process status has switched to the tapping period or the charging period, and the electrode has been pulled up to the highest position and locked; Alternatively, if the absolute value of the measured electrode speed is detected to be lower than the preset safe speed threshold and remains at the preset safe duration, the smooth switching module confirms that the safe switching conditions are met and starts the parameter writing process. During the writing process, the module constructs a transition coefficient that gradually changes from the corresponding value of the old parameter to the corresponding value of the parameter to be updated according to the preset transition time. The module performs linear interpolation of the old parameters in the main control parameter storage area to the parameter vector to be updated and writes the interpolation result into the main control area one scan cycle at a time. Each scan cycle uses the same set of real-time inputs to calculate the control results before and after the switching, and applies a limit to the proportional valve within the allowable drive voltage range for the interpolated parameters and control results, so that parameter changes and their corresponding control outputs remain continuous. Specifically, when the smooth switching module confirms that the safe switching conditions are met, it latches the old parameter vector, the parameter vector to be updated, and the transition start time, and divides the transition process into continuous control scan cycles according to the preset transition duration. Each scan cycle determines the current transition progress based on the number of cycles completed, and the progress increases monotonically from 0 to 1. Each item in the current master control parameter is linearly interpolated between the old parameter value and the parameter value to be updated according to the same progress. After the interpolated parameters are written into the main control parameter storage area, the main control execution module uses the electrode weight, hydraulic station pressure, furnace type identification and drive voltage collected in the same period to calculate the control results corresponding to the parameters before the switch and the control results corresponding to the parameters after the switch, and limits the allowable drive voltage range of the proportional valve for the measured output; after the transition is completed, the parameters to be updated are marked as the currently effective parameters and the old parameters and transition status are cleared; In each control cycle, the main control execution module calls the currently effective parameters in the main control parameter storage area to construct the basic proportional valve control signal. In order to cope with transient operating condition fluctuations, the execution module also obtains the real-time electrode weight deviation, hydraulic station pressure deviation, and the difference between the corresponding code value of the furnace type identification and the reference code value, and calculates the process correction amount by combining the correction coefficient in the storage area. Electrode weight deviation is the difference between the current electrode weight and the node average electrode weight; hydraulic station pressure deviation is the difference between the current pressure and the node average hydraulic station pressure; and the difference between the furnace type identifier corresponding code value and the reference code value is the difference between the current furnace type identifier corresponding code value and the reference code value. The three factors are multiplied by the corresponding correction coefficients pre-stored in the main control parameter storage area and then added together to form a process correction amount that corresponds one-to-one with the current operating condition. When any deviation item is missing, that item is treated as zero. This process correction amount is superimposed on the basic control signal, and after being processed by the system's amplitude limiting, it is finally output to the proportional valve drive end to form a closed-loop control. During the execution of parameter smooth switching, if a sudden abnormality occurs at the electric furnace site, causing a change in the process state, such as the operator unexpectedly triggering an emergency boost command at low speed, resulting in the safety speed condition no longer being met, the smooth switching module will immediately interrupt the transition interpolation process, freeze the currently constructed transition parameters as temporary master control parameters to respond to the emergency command, and after the system recovers to the safety switching condition, re-evaluate and resume the remaining interpolation transition process to ensure continuous control output. The temporary master control parameter only maintains the current interpolated parameter state and does not revert to the old parameter; when resuming switching, the transition progress at the previous interruption time is used as the starting point to continue interpolation, and the temporary master control parameter remains valid until the safe switching condition is detected again; In practice, the parameter verification module processes a batch of new parameters in the order of pending verification, verification passed, pending switching, switching in progress, switching paused and switching completed. The new fitting coefficients, node lookup table corrections, coverage, sample velocity standard deviation and root mean square error must belong to the same model scan and have a consistent parameter version identifier. If any data is missing, the version is inconsistent or the fitting interval is not fully participated in the evaluation, the status will directly return to pending verification and no parameters to be updated will be constructed. For coverage where larger index values ​​are better, or standard deviation and root mean square error where smaller index values ​​are better, the sub-item scores are first truncated within the corresponding qualified interval and then normalized. When the upper and lower qualified limits are valid, the parameter qualification is determined by summing the products of the three sub-item scores and their fixed weights. If any sub-item cannot be calculated, it will not be compensated by the other sub-items, and the parameters of this batch will be directly judged as unqualified. The smooth switching module determines the status according to the priority of safety conditions, in the following order: emergency control or over-limit protection, input data validity, process safety window, low-speed continuous condition, and normal parameter switching condition; when emergency control or proportional valve output exceeds the limit, the parameter progress increment is immediately stopped and the current transition parameter is frozen; when the input data is invalid, it is prohibited to continue writing new interpolation parameters. Resumption or switching is only permitted when both of the first two conditions are normal and one of the process safety window or low-speed continuous conditions is met; the safe speed duration is accumulated using continuous effective scan cycles, and the count is immediately reset to zero if the speed exceeds the threshold, the process status fails, or the electrode lockout is released; after switching is paused, the completed transition progress and a copy of the current parameters are retained, and the process continues from that progress when resuming, without restarting from the old parameters, to avoid reverse jumps in the control reference. Within each control scan cycle, the smooth switching module first reads the real-time input and current parameter version latched at the same moment, and then calculates the two sets of control results before and after the switch; only when both sets of results have completed the amplitude limiting, the parameter copy verification has passed, and the output version has not been updated by other tasks, will the interpolation parameters and control output of this cycle be submitted. If the submission conditions are not met, the parameters and outputs that have already taken effect in the previous cycle will be retained, and the switching status will be set to switching paused. After the transition progress reaches 1, the module will perform range checks and consistency checks on the parameters to be updated. Only after the checks pass will the parameters be marked as currently effective parameters. Otherwise, the pending update mark will be cleared and the safe parameters used during the switching pause will continue to be used.

[0021] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A linearization calculation system for proportional valves used in hydraulic lifting of electric furnaces, characterized in that, It includes a main controller and a main control parameter storage area, a data acquisition module, a steady-state judgment module, a data caching module, a parameter calculation module, a parameter verification module, a smooth switching module, and a main control execution module. The data acquisition module obtains the proportional valve drive voltage, measured electrode speed, hydraulic station pressure, electrode weight, furnace type identification, and process status. The main controller presets discrete voltage nodes within the driving voltage range of the proportional valve as reference voltage nodes. The steady-state judgment module outputs a sampling trigger signal when the proportional valve drive voltage is within the preset tolerance of the reference voltage node, and the rate of change of the hydraulic station pressure and the measured electrode speed are both no greater than the corresponding limit and continue for a preset time. The data caching module stores the acquired data as sampled data according to the corresponding reference voltage node based on the sampling trigger signal. When the coverage of the reference voltage nodes with the sampled data reaches a preset lower limit, the parameter calculation module establishes a fitting model to generate new parameters containing the control voltage correction amount corresponding to each reference voltage node. When the parameter verification module determines that an update is allowed, it marks the new parameter as a parameter to be updated. When the safety conditions are met, the smooth switching module writes the parameter to be updated into the main control parameter storage area. The main control execution module calls the parameters in the main control parameter storage area to generate a proportional valve control signal, and outputs it to the proportional valve drive end after superimposing the real-time calculated process correction amount.

2. The linearization calculation system for the proportional valve of the electric furnace hydraulic lifting according to claim 1, characterized in that, The data acquisition module obtains electrode displacement and measured electrode velocity from an external displacement detection device, hydraulic station pressure from an external hydraulic station pressure detection device, and electrode weight, furnace type identification and process status from an external host computer, manual input terminal or process database. The data acquisition module includes a proportional valve drive voltage acquisition unit, an electrode displacement acquisition unit, an electrode speed calculation unit, a hydraulic station pressure acquisition unit, a process status acquisition unit, a furnace type identification input unit, and an electrode weight input unit. The electrode displacement acquisition unit receives the displacement signal output by the displacement detection device, the hydraulic station pressure acquisition unit receives the pressure signal output by the hydraulic station pressure detection device, and the furnace type identification input unit, process status acquisition unit, and electrode weight input unit receive corresponding data provided by the host computer, manual input terminal, or process database.

3. The linearization calculation system for the proportional valve of the electric furnace hydraulic lifting according to claim 1, characterized in that, The steady-state judgment module calculates the pressure change rate of the hydraulic station and the measured electrode speed change rate according to a fixed sampling period; When the proportional valve drive voltage is within the preset tolerance of the reference voltage node, the hydraulic station pressure change rate is not greater than the preset pressure change rate limit, the measured electrode speed change rate is not greater than the preset speed change rate limit, and the above conditions are continuously met for a preset stable duration, the sampling trigger signal of the corresponding reference voltage node is output.

4. The linearization calculation system for the proportional valve of the electric furnace hydraulic lifting according to claim 3, characterized in that, The reference voltage node includes a range of -10V to +10V. The system includes 41 equally spaced nodes with a step size and including the endpoints. In addition to these 41 equally spaced nodes, a proportional valve dead zone node, obtained through a proportional valve no-response zone test, is added. The total number of reference voltage nodes is denoted as […]. ; The proportional valve dead zone node includes at least one of a forward opening voltage node, a reverse opening voltage node, and a zero-position node; the preset tolerance is... to The preset pressure change rate limit is to The preset speed change rate limit is to The preset stable duration is .

5. The linearization calculation system for the proportional valve of the electric furnace hydraulic lifting according to claim 1, characterized in that, The data caching module sets up an independent storage area for each reference voltage node. The independent storage area stores the node voltage identifier, the node average electrode speed, the node average electrode weight, the node average hydraulic station pressure, the furnace type identifier statistics, and the effective sample count. When the corresponding reference voltage node obtains sampling data, the node average electrode speed, node average electrode weight, and node average hydraulic station pressure are updated using an exponential moving average method. The furnace type identifier statistics are also updated based on the furnace type identifier that appears most frequently. The exponential moving average forgetting factor is [missing value]. to .

6. The linearization calculation system for the proportional valve of the electric furnace hydraulic lifting according to claim 5, characterized in that, The data caching module is further configured to eliminate abnormal data; when the difference between the measured electrode speed in the sampled data and the node average electrode speed of the corresponding reference voltage node exceeds the preset speed tolerance of the reference voltage node, the sampled data is not written into the corresponding independent storage area; the preset speed tolerance is determined according to the design target electrode speed and the speed detection precision, and the design target electrode speed is determined by the design motion curve pre-stored in the main control parameter storage area.

7. The linearization calculation system for the proportional valve of the electric furnace hydraulic lifting according to claim 1, characterized in that, The parameter calculation module calculates the effective sampling node coverage, where the effective sampling node coverage is the ratio of the number of nodes whose effective sample count reaches the minimum number of valid samples per node to the total number of reference voltage nodes; The minimum number of valid samples per node is 5 to 50, and the preset coverage lower limit is 0.6 to 0.9; when the effective sampling node coverage reaches the preset coverage lower limit, a multiple linear fitting model is established based on the driving voltage, electrode weight, hydraulic station pressure, the code value corresponding to the furnace type identifier and the measured electrode speed corresponding to the effective nodes, wherein the code value corresponding to the furnace type identifier is determined by the fixed coding table pre-stored in the main control parameter storage area.

8. The linearization calculation system for the proportional valve of the electric furnace hydraulic lifting according to claim 7, characterized in that, The parameter calculation module divides the voltage range from -10V to +10V into a high-speed descending zone, a low-speed descending zone, a dead zone and micro-movement zone, a low-speed ascending zone and a high-speed ascending zone according to preset voltage thresholds U1, U2, U3, U4, where U1 is the boundary threshold between the high-speed and low-speed descending zones, U2 is the boundary threshold between the low-speed descending zone and the dead zone, U3 is the boundary threshold between the dead zone and the low-speed ascending zone, U4 is the boundary threshold between the low-speed and high-speed ascending zones, and U1<U2<U3<U4, the threshold boundaries of each zone are determined according to the preset values pre-written in the main control parameter storage area; The maximum ascending speed and the maximum descending speed are process parameters pre-stored in the main control parameter storage area, and the design motion curve is a piecewise linear curve established according to the maximum ascending speed, the maximum descending speed and the driving voltage range.

9. The linearization calculation system for the proportional valve of the electric furnace hydraulic lifting according to claim 8, characterized in that, The parameter calculation module generates the control voltage correction amount of each reference voltage node according to the deviation between the design target electrode speed given by the design motion curve and the model predicted electrode speed output by the fitting model, wherein the design target electrode speed is determined by the design motion curve pre-stored in the main control parameter storage area; When the difference between the design target electrode speeds of adjacent nodes is not zero, the control voltage correction amount is generated by using the correction gain determined by the slope of the adjacent nodes, and when the difference between the design target electrode speeds of adjacent nodes is zero, the control voltage correction amount is generated by using a preset correction gain; the boundary nodes adopt a preset boundary correction gain; the main control execution module performs linear interpolation on the control voltage correction amount between adjacent reference voltage nodes according to the current driving voltage.

10. The linearization calculation system for the proportional valve of the electric furnace hydraulic lifting according to claim 1, characterized in that, The parameter verification module determines the parameter qualification degree according to the weighted evaluation result of the effective sampling node coverage, the sample speed standard deviation and the root mean square error. When the parameter qualification degree reaches the preset qualification threshold, the new fitting coefficient and the node look-up table correction amount output by the parameter calculation module are marked as parameters to be updated, and the preset qualification threshold is 0.7 to 0.95; The smooth switching module initiates writing when any of the following preset safe switching conditions are met: the process state is the steel tapping period or the feeding period and the electrode is in the highest locked state; or the measured absolute value of the electrode speed is not greater than the preset safe speed threshold and continues to reach the preset safe duration. The smooth switching module interpolates and switches the old parameter vector and the parameter vector to be updated according to a preset transition coefficient, and generates a proportional valve control output during the switching process. The process correction amount is determined based on the deviation between the current electrode weight and the historical average electrode weight, the deviation between the current hydraulic station pressure and the historical average hydraulic station pressure, and the difference between the corresponding code value of the furnace type identifier and the reference code value pre-stored in the main control parameter storage area. The relevant correction coefficients are stored in the main control parameter storage area, and the final output is limited to the allowable drive voltage range of the proportional valve.

Images