A method and system for risk identification monitoring and early warning of a silicon-manganese alloy production process

Abstract

This invention relates to the field of early warning technology in the metallurgical process of silicon-manganese alloys, and discloses a method and system for risk identification, monitoring, and early warning in the production process of silicon-manganese alloys. The method includes: acquiring electrode region operating data; constructing furnace condition health factors; performing steady-state determination of cotton placement; employing a multi-segment fuzzy interval division mechanism to execute energy consumption adjustment tasks in the curing section; constructing a pressure plate speed matching residual set; constructing a multi-source weighted health assessment model; and outputting production process risk early warning results. Compared to existing technologies that rely on single indicators such as current fluctuations or cotton placement stability for furnace condition judgment, especially in operating scenarios where furnace condition fluctuates or control parameters deviate from the set operating range, making it difficult to identify early furnace condition anomalies, this application achieves a quantitative assessment of the entire process health status based on furnace condition, electrodes, cotton placement, and speed residuals by constructing fuzzy membership mapping relationships between heterogeneous source indicators and introducing a nonlinear multi-parameter weighted fusion mechanism.

Description

Technical Field

[0001] This invention relates to the field of early warning technology for silicon-manganese alloy metallurgical processes, and in particular to a method and system for risk identification, monitoring and early warning in the production process of silicon-manganese alloys. Background Technology

[0002] Currently, silicon-manganese alloys are important alloying additives in the metallurgical industry. Their production process typically involves a carbothermic reduction reaction in a submerged arc furnace at high temperatures. This process is essentially a highly coupled electro-thermal-physical-chemical multi-physics system with a complex operating environment and numerous disturbance factors, including but not limited to electrode feeding and unloading states, changes in furnace charge structure, fluctuations in cooling water flow rate, fluctuations in hydraulic system, and electrode control errors. These factors can significantly affect furnace stability.

[0003] Current technologies for risk monitoring in the silicon-manganese alloy production process largely rely on low-frequency electrical parameter sampling, empirical threshold judgment, and manual inspection, lacking in-depth modeling and precise identification mechanisms for the "evolution of abnormal furnace conditions." For example, in the early stages of typical abnormal states such as slagging, collapse, ringing, flaming, or poor permeability, electrode current, voltage, and impedance signals often exhibit dynamic characteristics such as high-frequency perturbations, nonlinear drift, and asymmetric pulses. However, these abnormal signals are often submerged in normal fluctuations, making it difficult for traditional methods to identify their early signs. Furthermore, furnace condition fluctuations typically involve the coordinated interference of multiple subsystems, including electrical parameters, hydraulics, cooling, and feeding. Traditional methods have not constructed sufficient fusion sensing mechanisms, resulting in biased judgments and delayed responses. Additionally, cotton cloth, as a key buffer layer for the structural stability of the furnace charge column, plays a crucial role in regulating energy consumption and furnace conditions in the subsequent drying section. Traditional methods have not yet incorporated its dynamic behavior into the control decision-making closed loop.

[0004] Therefore, there is an urgent need for a risk identification, monitoring and early warning method for silicon-manganese alloy production process that can still achieve high-precision modeling and risk identification and early warning of furnace health status under complex disturbance backgrounds and variable operating conditions, so as to improve the safety, stability and automation level of the production process. Summary of the Invention

[0005] To address the aforementioned technical shortcomings, the purpose of this invention is to propose a risk identification, monitoring, and early warning method for the silicon-manganese alloy production process. This method aims to solve the technical problem in existing technologies that rely on single indicators such as current fluctuations or cotton cloth stability to judge furnace conditions, especially in operating scenarios where furnace conditions fluctuate or control parameters deviate from the set operating range, making it difficult to identify early abnormalities in furnace conditions.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The present invention provides a method for risk identification, monitoring and early warning in the production process of silicon-manganese alloy.

[0007] The method for risk identification, monitoring, and early warning in the silicon-manganese alloy production process includes:

[0008] Step S10: Obtain the electrode region operating data of the silicon-manganese alloy production furnace at time t. Based on the electrode region operating data, construct the furnace condition health factor using a three-dimensional impedance vector analysis mechanism driven by the energy imbalance index. ;

[0009] Step S20: Based on furnace health factors An asymmetric gated product machine with dynamic confidence modulation is used to perform the fabric stability judgment task and output the fabric steady-state confirmation index. ;

[0010] Step S30: Confirmation of steady-state indicators based on cotton fabric A multi-segment fuzzy interval division mechanism is used to perform the energy consumption adjustment task of the curing section, and the drying frequency control instruction set V is output. ;

[0011] Step S40: Control the drying frequency instruction set V The speed residual construction task is performed using a pressure plate speed synchronization feedback mechanism, and the pressure plate speed matching residual set is output. ;

[0012] Step S50: Finally, based on furnace health factors cotton steady-state confirmation indicators Matching residual set with platen speed A multi-source weighted health assessment model is constructed using a nonlinear weighting method based on fuzzy membership mapping. Multi-source weighted health assessment model Output risk warning results during the production process.

[0013] Preferably, in step S10, the electrode region operating data of the silicon-manganese alloy production furnace at time t is obtained, and a furnace health factor is constructed based on the electrode region operating data using a three-dimensional impedance vector analysis mechanism driven by the energy imbalance index. The steps specifically include:

[0014] Step S101: Obtain the electrode region operating data of the silicon-manganese alloy production furnace at time t. The electrode region operating data includes current I(t), voltage U(t), cooling water flow rate Q(t), and hydraulic pressure P(t). Introduce the current phase angle φ(t), and construct a complex impedance based on the current phase angle φ(t), current I(t), and voltage U(t). , Where j is the imaginary unit;

[0015] Step S102: Based on the complex impedance The complex plane angle index is calculated using the analytical method of cosine similarity of the angle between complex impedance vectors. Complex plane angle index This indicates whether the current impedance fluctuation tends to diverge or converge, corresponding to the conditions of collapsed furnace and electrode short-circuit furnace.

[0016] Step S103: Further based on the complex plane angle index The energy index of impedance change is defined using a complex conductivity energy analysis method driven by polar coordinate transformation. Impedance change energy index Used to indicate the degree of electrode impedance fluctuation;

[0017] Step S104: Energy index based on impedance change The cooling water flow rate Q(t) and hydraulic pressure P(t) are linearly fused using the Takagi-Sugeno type fuzzy inference principle to construct the final furnace health factor. .

[0018] Preferably, in step S102, the complex plane angle index The formula is expressed as:

[0019] ;

[0020] in, It is the inverse cosine function; This indicates taking the real part of a complex number; This indicates taking the imaginary part of a complex number; The preset sampling interval; For a moment The complex impedance.

[0021] Preferably, in step S103, the impedance change energy index The formula is expressed as:

[0022] ;

[0023] in, Indicates the instantaneous rate of change of the complex plane angle index; This represents the instantaneous rate of change of the complex impedance modulus.

[0024] Preferably, in step S20, based on furnace health factors... An asymmetric gated product machine with dynamic confidence modulation is used to perform the fabric stability judgment task and output the fabric steady-state confirmation index. The steps specifically include:

[0025] Step S201: Based on furnace health factors A confidence modulation factor is constructed using the Sigmoid modulation function, and the feature vector of the confidence modulation factor is output. ;

[0026] Step S202: Collect the cotton fabric section operation data, including the main fan frequency, upper layer cotton fabric speed, lower layer cotton fabric speed, and cotton collection motor speed; based on the cotton fabric section operation data, use an asymmetric gating function to extract dynamic gating factors of different dimensions, and output the dynamic gating factor feature vector. ;

[0027] Step S203: Based on the confidence modulation factor feature vector and dynamic gating factor eigenvector An asymmetric gated product aggregation method is used to perform fusion processing, outputting an asymmetric gated fusion vector. This asymmetric gated fusion vector is then subjected to gated nonlinear compression using a hyperbolic tangent function, ultimately outputting a fabric steady-state confirmation index. .

[0028] Preferably, in step S30, the steady-state confirmation index of the cotton fabric is used. A multi-segment fuzzy interval division mechanism is used to perform the energy consumption adjustment task of the curing section, and the drying frequency control instruction set V is output. The steps specifically include:

[0029] Step S301: Fuzzy Interval Mapping Stage: Using an exponential function to determine the steady-state confirmation index of cotton fabric. Mapped to multiple semantically fuzzy sub-intervals, including the first stable fuzzy sub-interval. Second stable fuzzy sub-interval and the third stable fuzzy sub-interval ;

[0030] Step S302: Construction stage of nonlinear cross-harmonic fusion function: Introducing the cotton stability trend gradient term , According to the gradient term of the stable trend of cotton fabric A nonlinear cross-harmonic fusion function is constructed using the gradient-weighted exponential harmonic mapping principle for multiple semantically fuzzy sub-intervals. This nonlinear cross-harmonic fusion function includes a first drying frequency adjustment function. Second drying frequency adjustment function and the third drying frequency adjustment function ;

[0031] Step S303: Based on the function value of the nonlinear cross-harmonic fusion function and combined with the preset dynamic mapping transformation rule, perform normalized linear proportional mapping and output the drying frequency control instruction set V. .

[0032] Preferably, in step S40, the drying frequency control instruction set V is used. The speed residual construction task is performed using a pressure plate speed synchronization feedback mechanism, and the pressure plate speed matching residual set is output. The steps specifically include:

[0033] Step S401: Based on the drying frequency control instruction set V The reference value of the target pressure plate speed was obtained by using the moving average method. ;

[0034] Step S402: Real-time acquisition of pressure plate operating parameters, including the current edge breaking machine speed. scraper running speed and pressure feedback value of the pressure plate cylinder Based on the current edge-breaking machine speed scraper running speed Pressure feedback value of the pressure plate cylinder Reference value of target pressure plate speed Construction pressure plate speed adjustment offset item ;

[0035] Step S403: Adjust the offset term based on the pressure plate speed Construct and output the pressure plate speed matching residual set .

[0036] This invention also provides a risk identification, monitoring, and early warning system for the silicon-manganese alloy production process, comprising:

[0037] The furnace health construction module is used to acquire the electrode region operating data of the silicon-manganese alloy production furnace at time t. Based on the electrode region operating data, a three-dimensional impedance vector analysis mechanism driven by the energy imbalance index is used to construct furnace health factors. ;

[0038] The cotton steady-state determination module is used to determine the health factors of the furnace condition. An asymmetric gated product machine with dynamic confidence modulation is used to perform the fabric stability judgment task and output the fabric steady-state confirmation index. ;

[0039] The drying frequency command generation module is used to confirm the steady-state performance of cotton fabric based on the cotton fabric steady-state indicator. A multi-segment fuzzy interval division mechanism is used to perform the energy consumption adjustment task of the curing section, and the drying frequency control instruction set V is output. ;

[0040] The platen residual construction module is used to control the instruction set V according to the drying frequency. The speed residual construction task is performed using a pressure plate speed synchronization feedback mechanism, and the pressure plate speed matching residual set is output. ;

[0041] The multi-source health assessment module is used for final assessment based on furnace condition health factors. cotton steady-state confirmation indicators Matching residual set with platen speed A multi-source weighted health assessment model is constructed using a nonlinear weighting method based on fuzzy membership mapping. Multi-source weighted health assessment model Output risk warning results during the production process.

[0042] The present invention also provides a risk identification, monitoring and early warning device for silicon-manganese alloy production process, comprising: a memory, a processor and a risk identification, monitoring and early warning program for silicon-manganese alloy production process stored in the memory and executable on the processor, wherein the risk identification, monitoring and early warning program for silicon-manganese alloy production process is executed by the processor to implement a risk identification, monitoring and early warning method for silicon-manganese alloy production process.

[0043] The present invention also provides a computer program product, including a risk identification, monitoring and early warning program for the silicon-manganese alloy production process, wherein the risk identification, monitoring and early warning program for the silicon-manganese alloy production process is executed by a processor to implement the aforementioned risk identification, monitoring and early warning method for the silicon-manganese alloy production process.

[0044] The beneficial effects of this invention are as follows: Based on the high-frequency waveform changes of electrode voltage, current and impedance, this invention can construct a furnace health factor that integrates the energy imbalance index and the three-dimensional impedance vector, thereby achieving highly sensitive capture of early characteristics of abnormal furnace conditions (such as slagging, collapse, ringing, fire, and poor permeability), and improving the timeliness and accuracy of risk identification.

[0045] This invention introduces an asymmetric gated product machine with dynamic confidence modulation and a fuzzy membership mapping weighting mechanism to construct a multi-source fusion model of cotton fabric steady state and pressure plate response. This enables comprehensive early warning of furnace instability trends under complex disturbance scenarios, effectively reducing the probability of abnormal evolution into serious failures and ensuring production continuity and safety. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is a flowchart illustrating the first embodiment of a method for risk identification, monitoring, and early warning in the production process of silicon-manganese alloy according to the present invention.

[0048] Figure 2 This is a schematic diagram comparing the three-dimensional impedance vector trajectories of the first embodiment of the method for risk identification, monitoring and early warning in the production process of silicon-manganese alloy according to the present invention.

[0049] Figure 3 This is a schematic diagram of the cotton stability identification results of the first embodiment of the risk identification, monitoring and early warning method for silicon-manganese alloy production process of the present invention.

[0050] Figure 4 This is a schematic diagram of the residual between the pressure plate target and the actual speed in the first embodiment of a risk identification, monitoring and early warning method for silicon-manganese alloy production process according to the present invention.

[0051] Figure 5 This is a schematic diagram of the influence distribution of the first health assessment model input for a first embodiment of a method for risk identification, monitoring and early warning in the production process of silicon-manganese alloy according to the present invention.

[0052] Figure 6 This is a schematic diagram of the influence distribution of the second health assessment model input for a first embodiment of a method for risk identification, monitoring and early warning in the production process of silicon-manganese alloy according to the present invention.

[0053] Figure 7 This is a schematic diagram of the influence distribution of the third health assessment model input in the first embodiment of the risk identification, monitoring and early warning method for silicon-manganese alloy production process of the present invention.

[0054] Figure 8 This is a schematic diagram of the equipment for a method of risk identification, monitoring and early warning in the production process of silicon-manganese alloy according to the present invention. Detailed Implementation

[0055] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] Example 1: As Figure 1 The diagram shown is a flowchart of the first embodiment of the method for risk identification, monitoring and early warning in the production process of silicon-manganese alloy according to the present invention. The first embodiment of the method for risk identification, monitoring and early warning in the production process of silicon-manganese alloy according to the present invention is presented.

[0057] In the first embodiment, the risk identification, monitoring, and early warning method for the silicon-manganese alloy production process includes:

[0058] Step S10: Obtain the electrode region operating data of the silicon-manganese alloy production furnace at time t. Based on the electrode region operating data, construct the furnace condition health factor using a three-dimensional impedance vector analysis mechanism driven by the energy imbalance index. ;

[0059] It should be noted that the "energy imbalance index-driven three-dimensional impedance vector analysis mechanism" refers to, after acquiring the electrode region operating data (including electrode voltage, current, interphase impedance, etc.) at the current sampling time t, constructing impedance vector models for each electrode in three-dimensional space, extracting their magnitude, phase angle, and relative distribution structure, and combining this with the energy input and output change trend per unit time, calculating the "energy imbalance index" representing the asymmetry of energy conversion and local abnormal response. This index is used to drive the weight adjustment process of the three-dimensional impedance vector model, thereby enhancing the sensitivity to subtle furnace condition disturbances. The operating data covers the three-phase electrode data channels, has millisecond-level resolution, and can accurately reflect the dynamic evolution trend of the furnace condition. Finally, the "furnace health factor" extracted through this mechanism provides a foundation for quantitatively characterizing the electrothermal coupling stability of the furnace body and identifying potential abnormal risks.

[0060] Understandably, by introducing a modulation mechanism based on the energy imbalance index, the nonlinear evolution characteristics of furnace conditions under high-frequency dynamic changes can be effectively identified. The construction of the three-dimensional impedance vector not only captures the changing trend of the traditional impedance modulus but also incorporates spatial geometric features such as phase angle shift and polar coordinate trajectory changes. This provides a more sensitive detection basis for issues such as changes in conductivity of the reaction zone inside the furnace, short-term thermal unevenness, and abnormal electrode propulsion. The output of this furnace condition health factor can serve as a leading indicator for subsequent steps in judging cotton stability, adjusting drying frequency, and generating risk warnings.

[0061] It should be understood that, compared to traditional furnace condition monitoring methods based on single-point impedance magnitude or voltage or current fluctuations, the three-dimensional impedance vector model constructed in this invention can achieve full spatial analysis of the electrode system's operating status. Furthermore, by actively adjusting feature extraction weights through an energy imbalance index, it overcomes signal masking problems caused by uneven electrode current load, changes in contact resistance, or sudden changes in the material structure of the submerged arc furnace. Under complex furnace conditions, this mechanism can proactively identify behaviors deviating from the normal thermal-electrical balance trajectory, thereby predicting furnace condition fluctuation trends in advance and providing earlier input for risk warning.

[0062] For example, such as Figure 2As shown, the evolution trend of the impedance vector trajectory of the three-phase electrode under normal and abnormal operating conditions is illustrated during the silicon-manganese alloy production process. It can be seen that under normal conditions, the vector trajectory distribution of the three-phase electrode is relatively balanced, showing good symmetry and small fluctuation amplitude, indicating that the furnace condition is in a stable operating range. Under abnormal conditions, such as slagging, material accumulation, or abnormal electrode contact, the vector trajectory shows obvious asymmetric stretching, phase angle shift, and violent amplitude fluctuations, reflecting the uneven energy distribution and local electrothermal coupling abnormalities.

[0063] Step S20: Based on furnace health factors An asymmetric gated product machine with dynamic confidence modulation is used to perform the fabric stability judgment task and output the fabric steady-state confirmation index. ;

[0064] It should be noted that the "asymmetric gated product machine with dynamic confidence modulation" refers to a process in which the furnace health factor constructed in the preceding steps is used as the core input variable during the cotton placement state identification process. A set of asymmetric gate functions (corresponding to the conduction zone, transition zone, and unstable zone in the cotton placement process, respectively) are constructed to capture the nonlinear joint activation effect between multidimensional furnace condition features through product coupling. Simultaneously, a dynamic confidence modulation mechanism is introduced to automatically adjust the sensitivity threshold and activation intensity of each gate function based on the stability of the furnace health factor, thereby adaptively adjusting the cotton placement judgment criteria at different stages of the process. The output of the cotton placement steady-state confirmation index is a set of scalars that can be used to indicate whether the current cotton placement process is in a stable, critical, or unstable state.

[0065] Understandably, this step, through multi-gated functions, models the piecewise response of furnace health factors, effectively distinguishing between fluctuations in cotton application caused by furnace disturbances and natural fluctuations during normal cotton application. Compared to single statistical indicators or fixed threshold judgment mechanisms, the asymmetric gated product machine models complex state boundary regions through the product coupling between gating functions, which helps to accurately identify minor anomalies in cotton application status. Simultaneously, the introduction of a dynamic confidence modulation mechanism gives the model adaptive capabilities, allowing it to update the judgment interval in real time based on the fluctuation amplitude of furnace health factors, avoiding over-triggered misjudgments.

[0066] For example, such as Figure 3As shown, gating function curves for the stable, transition, and unstable regions were constructed to address the variation of the furnace health factor value within the [0,1] interval. A dynamic confidence modulation factor was introduced in the transition region to enhance the sensitivity to furnace condition fluctuation boundaries. Finally, a "cotton fabric steady-state confirmation index" was constructed through the product of asymmetric gating functions. This index exhibits low response in both the stable and unstable regions, but high activation near the optimal state (moderate health factor), thus achieving high-precision identification and dynamic response to cotton fabric stability.

[0067] Step S30: Confirmation of steady-state indicators based on cotton fabric A multi-segment fuzzy interval division mechanism is used to perform the energy consumption adjustment task of the curing section, and the drying frequency control instruction set V is output. ;

[0068] It should be noted that the "multi-segment fuzzy interval division mechanism" refers to dividing the cotton fabric steady-state confirmation index output from the previous step S20 into continuous value intervals. During the division process, a fuzzy membership function is introduced to apply soft constraints to the interval boundaries, enabling the index to smoothly switch control strategies at interval transitions. Specifically, three core fuzzy intervals are preset: the over-dry warning zone (high steady state but high risk), the moderately dry zone (good steady state and reasonable energy consumption), and the under-dry response zone (insufficient steady state, requiring heating for enhancement). Each interval is defined by a Gaussian or triangular fuzzy function and weighted by combining the current thermal inertia coefficient of the furnace body with historical energy consumption rates, thereby determining parameters such as the target control frequency, adjustment rate, and heating power ratio in the drying frequency control instruction set V.

[0069] Understandably, this step uses a fuzzy interval division mechanism to continuously sense the stable state of the cotton fabric, so that control commands no longer rely on fixed threshold triggers, but can respond smoothly to slight fluctuations in stability indicators, avoiding frequent start-ups or frequency jumps that could impact the internal thermal field distribution of the furnace. Step S40: Control command set V according to the drying frequency. The speed residual construction task is performed using a pressure plate speed synchronization feedback mechanism, and the pressure plate speed matching residual set is output. ;

[0070] It should be noted that the "press plate speed synchronization feedback mechanism" refers to the dynamic setting of the target speed curve of the press plate during the curing stage, based on the drying frequency control command set V output in the preceding step S30, and the real-time acquisition of the actual operating speed of the press plate actuator, comparing it point by point through time synchronization. The core of this mechanism lies in constructing a set of speed matching residuals based on the difference between the target press plate speed and the actual press plate speed, thereby quantifying the response performance and dynamic deviation of the press plate control system. The speed residuals not only reflect the response delay of the control execution layer to high-frequency drying frequency commands, but also reveal inconsistencies in mechanical response caused by factors such as hydraulic system fluctuations, changes in press plate load, or local jamming.

[0071] It should be understood that traditional methods often focus on the achievement rate of the final displacement target, ignoring the details of the speed response during dynamic operation. Especially when the operating frequency of the pressure plate is controlled by the upper drying control logic, the target speed curve changes frequently, significantly increasing the requirements for transient response capability.

[0072] For example, such as Figure 4 As shown, the target pressure plate speed curve is set to 8, 12, 10, and 14 cm / s at different time periods. However, due to dynamic lag, execution deviation, and disturbance interference, the actual pressure plate response curve results in a non-stationary speed matching residual region between the two. The orange area in the figure intuitively reflects the dynamic evolution process of the "pressure plate speed matching residual set".

[0073] Step S50: Finally, based on furnace health factors cotton steady-state confirmation indicators Matching residual set with platen speed A multi-source weighted health assessment model is constructed using a nonlinear weighting method based on fuzzy membership mapping. Multi-source weighted health assessment model Output risk warning results during the production process.

[0074] It should be noted that the "nonlinear weighting method based on fuzzy membership mapping" refers to: constructing a family of fuzzy membership functions driven by domain experience parameters for each of the multi-source monitoring indicators obtained in the preceding steps, and setting different nonlinear weighting functions according to the influence sensitivity of each indicator to achieve dynamic weighted integration. The family of membership functions may include triangular, trapezoidal, Gaussian, etc., respectively mapping the degree of belonging of each indicator in multiple intervals such as "safe," "fluctuating," and "abnormal."

[0075] Understandably, by integrating fuzzy cognition and nonlinear weighting of multi-source key indicators, the accuracy and response speed of furnace condition risk identification can be significantly improved. Compared to traditional methods relying on single-indicator judgment or hard threshold classification, this method can effectively avoid problems such as boundary misjudgment and fluctuating alarms, and has a higher perception capability for furnace condition deterioration trends caused by multiple factors. During model construction, the fuzzy mapping mechanism can mitigate the interference of measurement errors and local fluctuations on the overall assessment results, while nonlinear weighting can focus on signal features with stronger influence, making the risk assessment more consistent with the dynamic evolution characteristics of the actual production process.

[0076] For example, such as Figure 5 As shown, with a residual of 0.05 (a proportional representation), the overall operation is stable, and the deviation of the pressure plate speed is minimal. At this point, the risk response surface mainly shows a monotonically increasing trend along the "furnace condition health factor" dimension, indicating a significant furnace condition-dominated effect, while changes in cotton fabric status have a limited impact on risk. High-risk areas are mainly concentrated in the upper right corner of the blast furnace condition imbalance area, and the overall risk distribution exhibits a "single-factor-dominated" structure, indicating that the model's judgment has high certainty. Figure 6 As shown, with a residual of 0.3, the fluctuations in the fabric and cotton state intensify, entering a relatively unstable range. The risk response surface expands from a unimodal to a bimodal structure, forming a new high-risk sensitive zone on the "fabric and cotton steady-state confirmation index" dimension, manifested as a significantly enhanced impact of thermal field shifts caused by fabric and cotton disturbances. The risk area expands to the central region, and the assessment model's responsiveness to multi-source disturbances improves, reflecting an approach to the unsteady-state boundary. Figure 7 As shown, when the residual is 0.7, the pressure plate speed control becomes significantly unbalanced, and the interaction between the three monitoring indicators intensifies. The risk response surface exhibits typical characteristics of multidimensional coupled disturbances, with the risk area spreading throughout the medium-to-high value space, the boundaries becoming increasingly blurred, and a more complex high-risk clustering pattern forming. At this point, the assessment model shows high sensitivity to changes in any indicator, possessing early warning capabilities, and enters an overall "multi-factor instability" state, exhibiting the strongest ability to capture abnormal trends.

[0077] Example 2: Furthermore, the present invention provides a risk identification, monitoring, and early warning system for a silicon-manganese alloy production process, employing a risk identification, monitoring, and early warning method for a silicon-manganese alloy production process as described in the above embodiments, which can solve the technical problem of risk identification, monitoring, and early warning in a silicon-manganese alloy production process. Compared with the prior art, the beneficial effects of the risk identification, monitoring, and early warning system for a silicon-manganese alloy production process provided by the present invention are the same as the beneficial effects of the risk identification, monitoring, and early warning method for a silicon-manganese alloy production process provided in the above embodiments, and other technical features of the risk identification, monitoring, and early warning system for a silicon-manganese alloy production process are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0078] Example 3: This invention provides a risk identification, monitoring, and early warning device for the silicon-manganese alloy production process. Please refer to... Figure 8 A risk identification, monitoring, and early warning device for a silicon-manganese alloy production process includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enable the at least one processor to perform the risk identification, monitoring, and early warning method for a silicon-manganese alloy production process described in Embodiment 1 above. The risk identification, monitoring, and early warning device for a silicon-manganese alloy production process in this embodiment of the invention may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), and vehicle terminals (e.g., vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. This risk identification, monitoring, and early warning device for a silicon-manganese alloy production process is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the invention. A risk identification, monitoring, and early warning device for a silicon-manganese alloy production process may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory 1002 or a program loaded from a storage device 1003 into a random access memory 1004. The random access memory 1004 also stores various programs and data required for the operation of the silicon-manganese alloy production process risk identification, monitoring, and early warning device. The processing unit 1001, the read-only memory 1002, and the random access memory 1004 are interconnected via a bus 1005. An I / O interface 1006 is also connected to the bus. Typically, the following systems can be connected to the I / O interface 1006: input devices 1007 including, for example, a touchscreen, touchpad, keyboard, mouse, image sensor, microphone, accelerometer, gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 1003 including, for example, magnetic tape, hard disk, etc.; and communication devices 1009. Communication device 1009 allows a silicon-manganese alloy production process risk identification, monitoring, and early warning device to wirelessly or wiredly communicate with other devices to exchange data. Although the figure shows a silicon-manganese alloy production process risk identification, monitoring, and early warning device with various systems, it should be understood that implementation or possession of all the systems shown is not required. More or fewer systems may be implemented alternatively.

[0079] Example 4: This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described method for risk identification, monitoring, and early warning in a silicon-manganese alloy production process. The computer program product provided by this invention can solve the technical problem of risk identification, monitoring, and early warning in a silicon-manganese alloy production process. Compared with the prior art, the beneficial effects of the computer program product provided by this invention are the same as those of the risk identification, monitoring, and early warning method for a silicon-manganese alloy production process provided in the above embodiments, and will not be repeated here.

[0080] In particular, according to the embodiments disclosed in this invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from read-only memory 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this invention.

[0081] It should be understood that the various parts disclosed in this invention can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0082] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for risk identification, monitoring, and early warning in the production process of silicon-manganese alloys, characterized in that, The methods include: Step S10: Obtain the electrode region operating data of the silicon-manganese alloy production furnace at time t. Based on the electrode region operating data, construct the furnace condition health factor using a three-dimensional impedance vector analysis mechanism driven by the energy imbalance index. Specifically, the electrode region operation data of the silicon-manganese alloy production furnace at time t is acquired, and a furnace health factor is constructed based on the electrode region operation data using a three-dimensional impedance vector analysis mechanism driven by the energy imbalance index. The steps specifically include: The electrode region operating data of the silicon-manganese alloy production furnace at time t is obtained. The electrode region operating data includes current I(t), voltage U(t), cooling water flow rate Q(t), and hydraulic pressure P(t). The current phase angle φ(t) is introduced, and a complex impedance is constructed based on the current phase angle φ(t), current I(t), and voltage U(t). , Where j is the imaginary unit; According to complex impedance The complex plane angle index is calculated using the analytical method of cosine similarity of the angle between complex impedance vectors. Complex plane angle index This indicates whether the current impedance fluctuation tends to diverge or converge, corresponding to the conditions of collapsed furnace and electrode short-circuit furnace. Further based on complex plane angle index The energy imbalance index is defined using a complex derivative energy analysis method driven by polar coordinate transformation. Energy Imbalance Index Used to indicate the degree of electrode impedance fluctuation; Based on the energy imbalance index The cooling water flow rate Q(t) and hydraulic pressure P(t) are linearly fused using the Takagi-Sugeno type fuzzy inference principle to construct the final furnace health factor. ; Step S20: Based on furnace health factors An asymmetric gated product machine with dynamic confidence modulation is used to perform the fabric stability judgment task and output the fabric steady-state confirmation index. ; Step S30: Confirmation of steady-state indicators based on cotton fabric A multi-segment fuzzy interval division mechanism is used to perform the energy consumption adjustment task of the curing section, and the drying frequency control instruction set V is output. ; Step S40: Control the drying frequency instruction set V The speed residual construction task is performed using a pressure plate speed synchronization feedback mechanism, and the pressure plate speed matching residual set is output. ; Step S50: Finally, based on furnace health factors cotton steady-state confirmation indicators Matching residual set with platen speed A multi-source weighted health assessment model is constructed using a nonlinear weighting method based on fuzzy membership mapping. Multi-source weighted health assessment model Output risk warning results during the production process.

2. The method for risk identification, monitoring, and early warning in the production process of silicon-manganese alloy as described in claim 1, characterized in that, In step S102, the complex plane angle index The formula is expressed as: ； in, It is the inverse cosine function; This indicates taking the real part of a complex number; This indicates taking the imaginary part of a complex number; The preset sampling interval; For a moment The complex impedance.

3. The method for risk identification, monitoring, and early warning in the production process of silicon-manganese alloy as described in claim 1, characterized in that, In step S103, the energy imbalance index The formula is expressed as: ； in, Indicates the instantaneous rate of change of the complex plane angle index; This represents the instantaneous rate of change of the complex impedance modulus.

4. The method for risk identification, monitoring, and early warning in the production process of silicon-manganese alloy as described in claim 1, characterized in that, In step S20, based on furnace health factors An asymmetric gated product machine with dynamic confidence modulation is used to perform the fabric stability judgment task and output the fabric steady-state confirmation index. The steps specifically include: Step S201: Based on furnace health factors A confidence modulation factor is constructed using the Sigmoid modulation function, and the feature vector of the confidence modulation factor is output. ; Step S202: Collect the cotton fabric section operation data, including the main fan frequency, upper layer cotton fabric speed, lower layer cotton fabric speed, and cotton collection motor speed; based on the cotton fabric section operation data, use an asymmetric gating function to extract dynamic gating factors of different dimensions, and output the dynamic gating factor feature vector. ; Step S203: Based on the confidence modulation factor feature vector and dynamic gating factor eigenvector An asymmetric gated product aggregation method is used to perform fusion processing, outputting an asymmetric gated fusion vector. This asymmetric gated fusion vector is then subjected to gated nonlinear compression using a hyperbolic tangent function, ultimately outputting a fabric steady-state confirmation index. .

5. The method for risk identification, monitoring, and early warning in the production process of silicon-manganese alloy as described in claim 1, characterized in that, In step S30, the steady-state confirmation index of cotton fabric is used. A multi-segment fuzzy interval division mechanism is used to perform the energy consumption adjustment task of the curing section, and the drying frequency control instruction set V is output. The steps specifically include: Step S301: Fuzzy Interval Mapping Stage: Using an exponential function to determine the steady-state confirmation index of cotton fabric. Mapped to multiple semantically fuzzy sub-intervals, including the first stable fuzzy sub-interval. Second stable fuzzy sub-interval and the third stable fuzzy sub-interval ; Step S302: Construction stage of nonlinear cross-harmonic fusion function: Introducing the cotton stability trend gradient term , According to the gradient term of the stable trend of cotton fabric A nonlinear cross-harmonic fusion function is constructed using the gradient-weighted exponential harmonic mapping principle for multiple semantically fuzzy sub-intervals. This nonlinear cross-harmonic fusion function includes a first drying frequency adjustment function. Second drying frequency adjustment function and the third drying frequency adjustment function ; Step S303: Based on the function value of the nonlinear cross-harmonic fusion function and combined with the preset dynamic mapping transformation rule, perform normalized linear proportional mapping and output the drying frequency control instruction set V. .

6. The method for risk identification, monitoring, and early warning in the production process of silicon-manganese alloy as described in claim 1, characterized in that, In step S40, according to the drying frequency control instruction set V The speed residual construction task is performed using a pressure plate speed synchronization feedback mechanism, and the pressure plate speed matching residual set is output. The steps specifically include: Step S401: Based on the drying frequency control instruction set V The reference value of the target pressure plate speed was obtained by using the moving average method. ; Step S402: Real-time acquisition of pressure plate operating parameters, including the current edge breaking machine speed. scraper running speed and pressure feedback value of the pressure plate cylinder Based on the current edge-breaking machine speed scraper running speed Pressure feedback value of the pressure plate cylinder Reference value of target pressure plate speed Construction pressure plate speed adjustment offset item ; Step S403: Adjust the offset term based on the pressure plate speed Construct and output the pressure plate speed matching residual set .

7. A risk identification, monitoring, and early warning system for a silicon-manganese alloy production process, applied to the risk identification, monitoring, and early warning method for a silicon-manganese alloy production process as described in any one of claims 1 to 6, characterized in that, The risk identification, monitoring, and early warning system for the silicon-manganese alloy production process includes: The furnace health construction module is used to acquire the electrode region operating data of the silicon-manganese alloy production furnace at time t. Based on the electrode region operating data, a three-dimensional impedance vector analysis mechanism driven by the energy imbalance index is used to construct furnace health factors. Specifically, the electrode region operation data of the silicon-manganese alloy production furnace at time t is acquired, and a furnace health factor is constructed based on the electrode region operation data using a three-dimensional impedance vector analysis mechanism driven by the energy imbalance index. The steps specifically include: The electrode region operating data of the silicon-manganese alloy production furnace at time t is obtained. The electrode region operating data includes current I(t), voltage U(t), cooling water flow rate Q(t), and hydraulic pressure P(t). The current phase angle φ(t) is introduced, and a complex impedance is constructed based on the current phase angle φ(t), current I(t), and voltage U(t). , Where j is the imaginary unit; According to complex impedance The complex plane angle index is calculated using the analytical method of cosine similarity of the angle between complex impedance vectors. Complex plane angle index This indicates whether the current impedance fluctuation tends to diverge or converge, corresponding to the conditions of collapsed furnace and electrode short-circuit furnace. Further based on complex plane angle index The energy imbalance index is defined using a complex derivative energy analysis method driven by polar coordinate transformation. Energy Imbalance Index Used to indicate the degree of electrode impedance fluctuation; Based on the energy imbalance index The cooling water flow rate Q(t) and hydraulic pressure P(t) are linearly fused using the Takagi-Sugeno type fuzzy inference principle to construct the final furnace health factor. ; The cotton steady-state determination module is used to determine the health factors of the furnace condition. An asymmetric gated product machine with dynamic confidence modulation is used to perform the fabric stability judgment task and output the fabric steady-state confirmation index. ; The drying frequency command generation module is used to confirm the steady-state performance of cotton fabric based on the cotton fabric steady-state indicator. A multi-segment fuzzy interval division mechanism is used to perform the energy consumption adjustment task of the curing section, and the drying frequency control instruction set V is output. ; The platen residual construction module is used to control the instruction set V according to the drying frequency. The speed residual construction task is performed using a pressure plate speed synchronization feedback mechanism, and the pressure plate speed matching residual set is output. ; The multi-source health assessment module is used for final assessment based on furnace condition health factors. cotton steady-state confirmation indicators Matching residual set with platen speed A multi-source weighted health assessment model is constructed using a nonlinear weighting method based on fuzzy membership mapping. Multi-source weighted health assessment model Output risk warning results during the production process.

8. A risk identification, monitoring, and early warning device for the silicon-manganese alloy production process, characterized in that, The risk identification, monitoring and early warning device for the silicon-manganese alloy production process includes: a memory, a processor, and a risk identification, monitoring and early warning program for the silicon-manganese alloy production process stored in the memory and executable on the processor. When the risk identification, monitoring and early warning program for the silicon-manganese alloy production process is executed by the processor, it implements a risk identification, monitoring and early warning method for the silicon-manganese alloy production process according to any one of claims 1 to 6.

9. A computer program product, characterized in that, The computer program product includes a risk identification, monitoring and early warning program for the silicon-manganese alloy production process. When the silicon-manganese alloy production process risk identification, monitoring and early warning program is executed by a processor, it implements a method for risk identification, monitoring and early warning of the silicon-manganese alloy production process according to any one of claims 1 to 6.

Images