Energy-saving control system of anode carbon block roasting furnace

Abstract

The invention relates to the technical field of anode carbon block roasting, and discloses an energy-saving control system of an anode carbon block roasting furnace, which comprises a data acquisition module, an energy-saving analysis module, a digital twin modeling module, a control execution module and a feedback optimization module, when the combustion state of the roasting furnace is monitored in real time, a multi-layer monitoring framework is constructed to dynamically track the change trend of temperature, flue gas oxygen content and pressure parameters, continuous comparison is performed based on a preset energy-saving reference interval, and when the parameters deviate from an optimization interval, a digital twinborn model is immediately triggered to intervene in analysis, so that the real-time monitoring of the combustion state of the roasting furnace is realized. Prejudgment identification of potential risks of combustion efficiency reduction and abnormal heat loss is achieved, hidden energy loss caused by imbalance of the air-fuel ratio is avoided, the non-planned shutdown frequency of equipment is reduced, and the continuous energy efficiency management and control stability of the production process is guaranteed.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of anode carbon block baking, in particular to an energy-saving control system of an anode carbon block baking furnace. BACKGROUND

[0002] The anode is the heart of the pre-baking tank aluminum electrolysis production, and is the main consumption material in the electrolytic aluminum production, participates in the reaction and is consumed in the electrochemical reaction of electrolysis, and the quality of the anode directly affects the stable production of the electrolysis tank and the yield and quality of the electrolysis and other technical and economic indexes. Anode baking is an important process of anode production in electrolytic aluminum, which aims to coking pitch, improving the mechanical properties of the anode, improving the good electrical conductivity and thermal conductivity of the anode, and reducing the oxidation reaction rate of the anode carbon block.

[0003] At present, due to the existence of multi-variable strong coupling characteristics in the operation process of the anode carbon block baking furnace, when the real-time regulation and control of the furnace combustion state is carried out, the fixed threshold control strategy equipped for the burner system cannot judge whether the actual air-fuel ratio in the combustion zone deviates from the optimal combustion interval in real time, and when local air-fuel ratio is out of control, it will lead to the decrease of combustion thermal efficiency, and cannot guarantee the optimization of energy utilization efficiency.

[0004] Therefore, the energy-saving control system of the anode carbon block baking furnace is proposed to solve the above problems. SUMMARY

[0005] In view of the deficiencies of the prior art, the present application provides an energy-saving control system of an anode carbon block baking furnace, which solves the problems raised in the background art.

[0006] In order to achieve the above purpose, the present application provides the following technical scheme: an energy-saving control system of an anode carbon block baking furnace, the system comprising: A data acquisition module configured to acquire real-time monitoring data by a distributed sensor network in real time, and output the real-time monitoring data; An energy-saving analysis module configured to receive the real-time monitoring data, determine whether the key parameters deviate from the preset energy-saving interval through a double-threshold judgment mechanism, and output a model activation instruction; A digital twin modeling module configured to respond to the model activation instruction, construct a virtual mapping of the furnace based on a historical database and real-time monitoring data, generate an energy consumption state evolution track and an optimization path deviation early warning, and output a prediction result and a risk level; A control execution module configured to receive the prediction result and the risk level, generate a dynamic adjustment instruction in combination with the current production task and the furnace health tolerance, and output the dynamic adjustment instruction to a burner execution mechanism to adjust the air-coal ratio and the gas supply amount; The feedback optimization module is configured to collect the post-execution kiln state data, verify the control effect through the energy consumption change rate mapping matrix, generate a control strategy correction instruction and feed back to the control execution module.

[0007] Preferably, the process of judging whether the digital twin model activation threshold is reached is: Set key energy-saving parameters, including real-time temperature value , energy-saving alert value and digital twin model activation threshold Calculate the deviation of the current parameter relative to the energy-saving interval, the formula is: ; When , it means that the parameter has exceeded the alert value, and the activation approach value is further calculated: ; Wherein is the temperature margin difference, is the maximum allowed temperature, is the actual temperature; When the activation approach value is less than the preset activation approach threshold , the digital twin model is triggered to make state prediction, otherwise the regular monitoring is maintained.

[0008] Preferably, the process of generating the current energy consumption state evolution trajectory is: Based on the output of the data acquisition module, a multi-dimensional time sequence curve of the kiln state is constructed with the X-axis as time and the Y-axis as each energy consumption parameter; Call the standard energy-saving trajectory curve in the historical database as the reference curve; Select the sampling points with the same time points on the current state curve and the reference curve, integrate the current curve parameter values to form the first state data group, and integrate the reference curve parameter values to form the second state data group; Calculate the cosine similarity between the two groups of data as the trajectory similarity value, the formula is: ; Wherein and are the parameter values of the first and second state data groups respectively, is the cosine similarity, is the vector dimension, is the index variable; Set the curve similarity threshold , when the similarity value is less than the threshold, it is determined that the curve type is linear, otherwise it is nonlinear, and the subsequent trend prediction is performed.

[0009] Preferably, the process of predicting whether the energy consumption trend deviates from the optimization path includes: When the curve type is linear, the least square method is used to fit the trend of the change of the key parameters to obtain a linear equation: ; wherein is the parameter value, is the time, is the slope, is the intercept; The energy consumption failure threshold is input into the equation to predict the time point at which the parameter reaches the threshold, i.e. the critical time point. When the curve type is nonlinear, the first state data set is subjected to a difference operation to calculate the maximum growth rate of the parameter, and based on the difference between the current value and the failure threshold, the critical duration is calculated: ; The critical time point is evaluated for energy consumption risk. If it is earlier than the end time point of the maintenance cycle, it is determined that there is a deviation risk.

[0010] Preferably, the process of evaluating the potential energy waste risk is: By matching the state trajectory change mode with the historical failure library, common waste paths, including insufficient combustion and heat loss, are identified; According to the trajectory similarity and the critical time point, the waste risk level is calculated: high, medium and low, and the expected energy consumption loss is quantified; When the risk level is higher than the tolerance threshold, the control execution module is entered, otherwise only the log is recorded.

[0011] Preferably, the process of determining whether intervention is needed in combination with the current production task and the furnace health tolerance is: Obtain production plan information and furnace health tolerance, the latter being defined as the difference between the current health score and the lower limit of tolerance; Construct a device health data set: including residual life, failure probability and production task data set: including load demand, urgency; The correlation between the two sets of data is calculated using the Pearson correlation coefficient, and the formula is: ; wherein is the Pearson correlation coefficient, and are the observed values, , are the mean values; When the correlation is high and the health tolerance is close to the threshold, it is determined that intervention is needed, otherwise the processing is delayed.

[0012] Preferably, the process of generating energy-saving control instructions is: Based on the feedback of the digital twin model, combined with the real-time working condition, the intervention strategy library is matched, the adaptive measures are screened, including load reduction operation, wind-coal ratio optimization and local shutdown, and according to the resource scheduling: spare parts availability, personnel configuration and production task, a detailed scheme is generated, including intervention content, time window and expected energy saving effect.

[0013] Preferably, the adaptive mechanism for executing the energy-saving control instruction is: The control execution module outputs signals to the burner actuator to dynamically adjust the gas valve and the fan speed; Through the feedback optimization module, the adjusted parameters are monitored in real time, and when the deviation persists, the digital twin model is reactivated for iterative optimization; The system has built-in fault tolerance logic, which automatically switches to backup strategies when external interference occurs.

[0014] Preferably, the sensor network of the data acquisition module includes: Infrared temperature sensor, pressure transmitter, gas analyzer and environmental temperature and humidity probe, deployed in key positions of the furnace, including the combustion zone and the exhaust port; Data is transmitted through industrial Ethernet, with a sampling frequency higher than 1Hz; The module integrates a data preprocessing unit to compensate for environmental interference.

[0015] Preferably, the historical database of the digital twin modeling module is constructed in the following way: Store the historical furnace operation data, including temperature curve, energy consumption record, fault event and maintenance log; The database uses a time series compression algorithm to reduce storage overhead and supports real-time updates through an API interface; Model training is based on a deep learning framework and is periodically optimized to adapt to new working conditions.

[0016] Compared with the prior art, the energy-saving control system for the anode carbon block roasting furnace provided by the present application has the following beneficial effects: 1. In the present application, when the roasting furnace combustion state is monitored in real time, the multi-level monitoring architecture is constructed to dynamically track the temperature, oxygen content in flue gas and pressure parameter change trend, and based on the preset energy-saving reference interval, the digital twin model is triggered for intervention analysis when the parameters deviate from the optimization interval, realizing the predictive identification of potential risks such as combustion efficiency decline and abnormal heat loss, avoiding hidden energy loss due to air-fuel ratio imbalance, reducing the frequency of unplanned equipment downtime, and ensuring the continuous performance of the production process and the stability of the control system.

[0017] 2. In the present application, in the process of executing the kiln dynamic regulation, by fusing the equipment health tolerance evaluation and the real-time production task demand, the running condition constraint condition is autonomously analyzed, the energy-saving regulation strategy matched with the current roasting stage is generated, and the key parameters of the burner air-coal ratio and the gas supply amount are implemented closed-loop self-correction based on the feedback data after execution, so that the system can adaptively relieve the temperature field disturbance caused by uneven material distribution and load mutation, and at the same time, when external environmental interference occurs, the standby control strategy is intelligently switched, to ensure the execution accuracy and system robustness of the control instruction under complex working conditions.

[0018] 3. In the present application, in the multi-kiln collaborative operation scene, by dynamically partitioning modeling the energy consumption evolution track of each independent roasting furnace, combining with the deep analysis of the differential performance consumption mode of the historical fault feature library, and generating a ladder type optimization strategy based on the partitioned energy consumption state, the coordination of gas source distribution and combustion parameters between the furnace groups is realized, the chain performance imbalance caused by gas source contention in traditional control is blocked, and the regulation strength is customized and adapted to the kiln with different health states according to the real-time working condition, to improve the balance of the overall energy efficiency situation of the roasting workshop and the intelligent level of production scheduling. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 The figure is the framework of the energy-saving control system of the anode carbon block roasting furnace. DETAILED DESCRIPTION

[0020] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0021] Specific embodiment: please refer to Figure 1 , the energy-saving control system of the anode carbon block roasting furnace, the system comprises: The data acquisition module is configured to acquire the temperature, oxygen content in flue gas, pressure and environmental parameters in the furnace in real time through a distributed sensor network, and output real-time monitoring data; The energy-saving analysis module is configured to receive the real-time monitoring data, determine whether the key parameters deviate from the preset energy-saving interval through a double-threshold judgment mechanism, and output a model activation instruction; The digital twin modeling module is configured to respond to the model activation instruction, construct a virtual mapping of the kiln based on the historical database and real-time monitoring data, generate an energy consumption state evolution track and an optimization path deviation warning, and output a prediction result and a risk level; The control execution module is configured to receive the prediction result and the risk level, combine the current production task and the kiln health tolerance to generate a dynamic adjustment instruction, and output the dynamic adjustment instruction to a burner execution mechanism to adjust the air-coal ratio and the air supply amount; The feedback optimization module is configured to collect kiln state data after execution, verify the control effect through an energy consumption change rate mapping matrix, generate a control strategy correction instruction, and feed back the control strategy correction instruction to the control execution module.

[0022] The process of determining whether the digital twin model activation threshold is reached is: Set key energy-saving parameters, including real-time temperature value , energy-saving alert value , and digital twin model activation threshold Calculate the deviation of the current parameter relative to the energy-saving interval, the formula is: ; When , it means that the parameter has exceeded the alert value, and the activation proximity value is further calculated: ; Wherein is the temperature margin difference, is the maximum allowed temperature, is the actual temperature; When the activation proximity value is less than the preset activation proximity threshold , the digital twin model is triggered to perform state prediction, otherwise the regular monitoring is maintained.

[0023] The process of generating the current energy consumption state evolution track is: Based on the output of the data acquisition module, a kiln state multi-dimensional time sequence curve is constructed with the X-axis as time and the Y-axis as each energy consumption parameter; Call the standard energy-saving track curve in the historical database as a reference curve; Select sampling points with the same time points on the current state curve and the reference curve, integrate the current curve parameter values to form a first state data group, and integrate the reference curve parameter values to form a second state data group; Calculate the cosine similarity between the two groups of data as the track similarity value, the formula is: ; Wherein and are the parameter values of the first and second state data groups respectively, is the cosine similarity, is the vector dimension, is the index variable; Set the curve similarity threshold When the similarity value is less than the threshold value, the curve type is determined to be linear, otherwise nonlinear, and subsequent trend prediction is performed.

[0024] The process of predicting whether the energy consumption trend deviates from the optimization path includes: When the curve type is linear, the least squares method is used to fit the trend of the change in the key parameters to obtain a straight line equation: ; wherein is the parameter value, is the time, is the slope, is the intercept; The energy consumption failure threshold value is input into the equation to predict the time point at which the parameter reaches the threshold value, i.e., the critical time point. When the curve type is nonlinear, a difference operation is performed on the first state data set to calculate the maximum growth rate of the parameter, and based on the difference between the current value and the failure threshold value, the critical duration is calculated: ; The critical time point is evaluated for energy consumption risk, and if it is earlier than the end time point of the maintenance cycle, it is determined that there is a deviation risk.

[0025] The process of evaluating potential energy waste risk is: By matching the state trajectory change pattern with the historical failure library, common waste paths, including insufficient combustion and heat loss, are identified; According to the trajectory similarity and the critical time point, the waste risk level is calculated: high, medium, and low, and the expected energy consumption loss is quantified. The calculation formula of the combustion efficiency loss rate is: ; wherein is the real-time combustion efficiency loss rate, is the measured air-fuel ratio, is the optimized air-fuel ratio, is the measured temperature, is the optimized temperature, is the maximum temperature deviation allowed by the system; When is greater than the preset loss threshold value, it is determined that the risk level is high, and the control execution module is entered, otherwise only the log is recorded.

[0026] The process of determining whether intervention is needed in combination with the current production task and the kiln health tolerance is: The production plan information and the kiln health tolerance are obtained, and the latter is calculated by the following formula to obtain the comprehensive health score: ; wherein is the comprehensive health score, The remaining life of the furnace, Design life, Real-time failure probability, The current capacity, The maximum design capacity, The weight coefficient, satisfying The default values are 0.4, 0.4, and 0.2, respectively. Constructing the device health data set: including And the remaining life and production task data set: including load demand, urgency; Calculate the correlation between the two sets of data using Pearson correlation coefficient, the formula is: ; Where The Pearson correlation coefficient, And The observed value, , The mean; When the correlation is high and the health tolerance is close to the threshold, it is determined that intervention is needed, otherwise delayed processing.

[0027] The process of generating energy-saving control instructions is: Based on the feedback of the digital twin model, combined with the real-time working condition, match the intervention strategy library, screen the adaptive measures, including load reduction operation, wind-coal ratio optimization and local shutdown, according to the resource scheduling: spare parts availability, personnel configuration and production task to generate detailed scheme, including intervention content, time window and expected energy-saving effect.

[0028] The adaptive mechanism for executing energy-saving control instructions is: The control execution module outputs signals to the burner actuator, dynamically adjusts the air supply valve and fan speed; Through the feedback optimization module, real-time monitoring of the adjusted parameters, when the deviation persists, re-activate the digital twin model iterative optimization, the specific control compensation amount is calculated by the following formula: ; Where The dynamic compensation of the actuator, Real-time control deviation, The proportional, integral, and differential coefficients, The differential symbol; The system has built-in fault tolerance logic, when external interference occurs, automatically switch to the backup strategy.

[0029] The sensor network of the data acquisition module includes: Infrared temperature sensor, pressure transmitter, gas analyzer and environmental temperature and humidity probe, deployed in key positions of the furnace, including the combustion zone and the exhaust port; Data is transmitted through industrial Ethernet, with a sampling frequency higher than 1Hz; The module integrates a data preprocessing unit to compensate for environmental interference, and the compensation coefficient formula is: ; Wherein The environmental correction factor, is the measured environmental humidity, is the reference humidity, is the maximum allowed humidity deviation, is the real-time atmospheric pressure, is the standard atmospheric pressure; The compensated data is analyzed subsequently to improve the analysis accuracy.

[0030] The historical database of the digital twin modeling module is constructed in the following way: Store the running data of the furnace in previous years, including temperature curve, energy consumption record, fault event and maintenance log; The database uses a time series compression algorithm to reduce storage overhead and supports real-time updates through an API interface; Model training is based on a deep learning framework, and the prediction error rate is verified by the following formula: ; Wherein is the normalized prediction error rate, is the predicted value at the ith time point, is the corresponding measured value, is the range of the parameter, n is the total number of samples, and i is the index variable; The model is periodically optimized to adapt to new working conditions.

[0031] The operation steps of the energy-saving control system of the anode carbon block roasting furnace are as follows: Step 1: Real-time acquisition and preprocessing of multi-source data The system continuously acquires temperature, oxygen content in flue gas, pressure parameters and environmental temperature and humidity data in the furnace through a distributed sensor array, and uses a data preprocessing unit to compensate for environmental interference and filter noise. The sensor network transmits standardized data through industrial Ethernet to the central processor at a sampling frequency higher than 1Hz, and builds a real-time monitoring database to provide a high-precision data foundation for subsequent analysis.

[0032] Step 2: Energy-saving state threshold determination and model activation The energy-saving analysis module dynamically compares real-time parameters with preset energy-saving intervals. When a key parameter deviates from the warning value, the deviation is calculated and it is assessed whether the digital twin model activation threshold is reached. When the activation condition is met, the virtual mapping construction process is triggered, otherwise the regular monitoring state is maintained. This decision mechanism avoids over-response to slight fluctuations and ensures efficient use of system resources.

[0033] Step three, energy consumption evolution trajectory modeling and risk prediction The digital twin modeling module constructs a virtual mapping of the furnace operation based on the historical database and real-time data, generating a multi-dimensional time series curve of the current energy consumption state. By calculating the cosine similarity between this curve and the standard energy-saving trajectory, the trajectory type is identified as linear or nonlinear, and the critical time point when the key parameter reaches the energy consumption failure threshold is predicted. Combined with historical fault library feature matching, the energy waste risk level and potential loss amount are quantified.

[0034] Step four, dynamic control instruction generation and execution The control execution module integrates risk prediction results, device health tolerance score, and real-time production task requirements to select appropriate measures from the intervention strategy library, including wind-coal ratio optimization, local cooling, and load adjustment. A detailed plan containing execution content, time window, and expected energy-saving effect is generated, and control signals are output to the burner actuator to dynamically adjust the air valve opening and fan speed, achieving control of the combustion process.

[0035] Step five, closed-loop feedback and strategy iteration optimization The feedback optimization module collects real-time data on the changes in the furnace state after execution and verifies the control effect through the energy consumption change rate mapping matrix. When it detects that parameter deviations persist, it reactivates the digital twin model for trajectory correction analysis, generates control parameter compensation instructions, and feeds them back to the control execution module. The system has a built-in fault tolerance mechanism that automatically switches to a backup strategy when external interference occurs, ensuring continuous control.

[0036] Step six, multi-furnace coordinated energy efficiency balancing For multi-furnace parallel scenarios in the roasting workshop, the system independently partitions and analyzes the energy consumption trajectories of each furnace, identifying different performance consumption patterns. Based on the real-time working conditions and health status of each furnace, a stepwise optimization strategy is developed and the gas source distribution priority is coordinated, balancing the energy supply and demand among the furnace group, blocking the gas source competition phenomenon in traditional control, and achieving overall energy efficiency improvement at the workshop level.

[0037] Step seven, model self-evolution and knowledge base update The historical database continuously stores the running data and the regulation results, and the digital twin model is periodically trained through a deep learning framework. The model accuracy is verified by using the prediction error rate, and when the error exceeds the threshold, the parameter optimization process is started. New fault events and solutions are simultaneously entered into the knowledge base, so that the system has the evolution ability to continuously adapt to new working conditions, forming an intelligent closed loop of "data accumulation-model optimization-control upgrade".

[0038] It should be noted that the relational terms herein such as first and second and the like are used solely to distinguish one entity or action from another, without necessarily requiring or implying any such actual relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element preceded by "comprises a" does not, without more limitations, foreclose the existence of additional identical elements in the process, method, article, or apparatus that comprises the recited element.

[0039] Although embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, alternatives, and variations can be made in the embodiments without departing from the spirit and scope of the present application as defined by the appended claims and their equivalents.

Claims

1. An energy-saving control system for an anode carbon block roasting furnace, characterized in that: The system includes: The data acquisition module is configured to collect furnace temperature, flue gas oxygen content, pressure and environmental parameters in real time through a distributed sensor network, and output real-time monitoring data. The energy-saving analysis module is configured to receive the real-time monitoring data, determine whether key parameters deviate from the preset energy-saving range through a dual threshold determination mechanism, and output a model activation command. The digital twin modeling module is configured to respond to the model activation command, construct a virtual mapping of the furnace based on historical database and real-time monitoring data, generate an energy consumption state evolution trajectory and an early warning of deviation from the optimization path, and output prediction results and risk levels. The control execution module is configured to receive the prediction results and risk level, combine them with the current production task and the furnace health tolerance to generate dynamic adjustment instructions, and output them to the burner actuator to adjust the air-coal ratio and gas supply. The feedback optimization module is configured to collect furnace status data after execution, verify the control effect through the energy consumption change rate mapping matrix, generate control strategy correction instructions, and feed them back to the control execution module.

2. The energy-saving control system for the anode carbon block roasting furnace according to claim 1, characterized in that: The process of determining whether the activation threshold of the digital twin model has been reached is as follows: Set key energy-saving parameters, including real-time temperature values. Energy saving warning value Activation threshold of digital twin model The formula for calculating the deviation of the current parameter from the energy-saving range is as follows: ； when This indicates that the parameter has exceeded the warning value, and further calculation of the activation proximity value is needed: ； in Due to temperature margin difference, For the maximum allowable temperature, This refers to the actual temperature. When the activation proximity value is less than the preset activation proximity threshold If the condition is met, the digital twin model will be triggered to predict the state; otherwise, regular monitoring will continue.

3. The energy-saving control system for the anode carbon block roasting furnace according to claim 1, characterized in that: The process of generating the current energy consumption state evolution trajectory is as follows: Based on the output of the data acquisition module, a multi-dimensional time-series curve of the furnace state is constructed with the X-axis representing time and the Y-axis representing each energy consumption parameter; Use the standard energy-saving trajectory curve from the historical database as a reference curve; Select sampling points with the same time point on the current state curve and the reference curve, integrate the parameter values ​​of the current curve to form the first state data group, and the parameter values ​​of the reference curve to form the second state data group; The cosine similarity between two sets of data is calculated as the trajectory similarity value, using the following formula: ； in and These are the parameter values ​​for the first and second state data groups, respectively. For cosine similarity, For vector dimensions, For index variables; Set curve similarity threshold If the similarity value is less than the threshold, the curve type is determined to be linear; otherwise, it is nonlinear, and subsequent trend prediction is performed.

4. The energy-saving control system for the anode carbon block roasting furnace according to claim 3, characterized in that: The process of predicting whether energy consumption trends deviate from the optimal path includes: When the curve is linear, the least squares method is used to fit the changing trend of the key parameters to obtain the equation of the straight line: ； in For parameter values, For time, The slope The intercept; The energy consumption fault threshold is input into the equation to predict the time point when the parameter reaches the threshold, i.e., the critical time point. When the curve type is nonlinear, a difference operation is performed on the first state data set to calculate the maximum growth rate of the parameter, and the critical duration is calculated based on the difference between the current value and the fault threshold. ； Energy consumption risk is assessed at this critical time point, and if it is earlier than the end of the maintenance cycle, it is determined that there is a deviation risk.

5. The energy-saving control system for the anode carbon block roasting furnace according to claim 1, characterized in that: The process of assessing the risk of potential energy waste is as follows: By matching state trajectory change patterns with historical fault databases, common waste paths are identified, including incomplete combustion and heat loss. Based on trajectory similarity and critical time points, the waste risk level is calculated as high, medium, or low, and the expected energy loss is quantified. If the risk level is higher than the tolerance threshold, the system will enter the control execution module; otherwise, it will only log the information.

6. The energy-saving control system for the anode carbon block roasting furnace according to claim 1, characterized in that: The process of determining whether intervention is needed based on current production tasks and the furnace health tolerance is as follows: Obtain production planning information and furnace health tolerance, the latter defined as the difference between the current health score and the tolerance lower limit; Build equipment health data sets: including remaining lifespan and failure probability, and production task data sets: including load demand and urgency. The correlation between two sets of data can be calculated using the Pearson correlation coefficient, as shown in the formula: ； in The Pearson correlation coefficient is used. and For the observed values, , The mean; If the correlation is high and the health tolerance is close to the threshold, then intervention is deemed necessary; otherwise, treatment is delayed.

7. The energy-saving control system for the anode carbon block roasting furnace according to claim 1, characterized in that: The process of generating energy-saving control instructions is as follows: Based on the control parameters fed back by the digital twin model, combined with real-time operating conditions, the intervention strategy library is matched and suitable measures are selected, including load reduction operation, wind-coal ratio optimization and partial shutdown. According to resource scheduling: spare parts availability, personnel configuration and production tasks, detailed plans are generated, including intervention content, time window and expected energy saving effect.

8. The energy-saving control system for the anode carbon block roasting furnace according to claim 7, characterized in that: The adaptive mechanism for executing energy-saving control commands is as follows: The control execution module outputs signals to the burner actuator to dynamically adjust the gas supply valve and fan speed; The feedback optimization module monitors the adjusted parameters in real time, and reactivates the digital twin model for iterative optimization if the deviation persists. The system has built-in fault-tolerant logic that automatically switches to a backup strategy when external interference occurs.

9. The energy-saving control system for the anode carbon block roasting furnace according to claim 1, characterized in that: The sensor network of the data acquisition module includes: Infrared temperature sensors, pressure transmitters, gas analyzers, and ambient temperature and humidity probes are deployed in key locations of the furnace, including the combustion zone and flue gas outlets. Data is transmitted via industrial Ethernet with a sampling frequency higher than 1Hz; The module integrates a data preprocessing unit to perform compensation calculations for environmental interference.

10. The energy-saving control system for the anode carbon block roasting furnace according to claim 1, characterized in that: The historical database construction method for the digital twin modeling module is as follows: Stores historical furnace operation data, including temperature curves, energy consumption records, fault events, and maintenance logs; The database uses a time-series compression algorithm to reduce storage overhead and supports real-time updates through an API interface. The model is trained based on a deep learning framework and is regularly optimized to adapt to new working conditions.

Images