Calcium carbide furnace pressure intelligent operation control method

Abstract

The invention belongs to the field of calcium carbide furnace pressure intelligent operation control, and particularly relates to a calcium carbide furnace pressure intelligent operation control method which comprises the following steps: continuously collecting operation process data of a calcium carbide furnace group and preprocessing; calculating by adopting trend analysis and short-term load prediction; the furnace condition operation parameters and the alarm records are analyzed, and the furnace condition health index of each calcium carbide furnace is calculated; the target total furnace pressure and the total furnace gas amount are decomposed, the load sharing coefficient of each calcium carbide furnace is calculated, an adjusting instruction of each calcium carbide furnace is generated, and cooperative adjustment is carried out on the calcium carbide furnaces of the furnace group; the deviation between the predicted gas consumption and the actual gas consumption and the furnace condition health index change are monitored in real time, and the gas consumption prediction model and the load sharing coefficient are dynamically adjusted through self-adaptive correction and abnormal rollback; according to gas consumption demand prediction, furnace condition health assessment and load sharing calculation based on furnace condition health data, the furnace pressure of the multiple calcium carbide furnaces can be cooperatively adjusted, and therefore stable operation of the furnace pressure is achieved.

Description

Technical Field

[0001] This invention belongs to the field of intelligent operation control of calcium carbide furnace pressure, and specifically relates to an intelligent operation control method for calcium carbide furnace pressure. Background Technology

[0002] The Chinese patent application (CN202010782420.5) discloses an intelligent operation control method for calcium carbide furnace pressure. This method sets the range of pressure variation and the standard pressure value for normal operation of the calcium carbide furnace, calculates the deviation and the rate of change of deviation; sets pressure limits for the calcium carbide furnace, collects the furnace pressure signal in real time, and preprocesses the signal; analyzes the pressure changes, performs fuzzy inference using the limit shifting method, selects the appropriate control strategy, and adjusts the furnace pressure by driving the exhaust fan through the output module.

[0003] Although this invention can respond promptly to changes in furnace pressure and control the exhaust fan to achieve rapid compensation and correction in the field of automatic control technology for calcium carbide furnaces, existing automatic furnace pressure control technologies mostly rely solely on the furnace pressure signal of a single calcium carbide furnace as the only control basis. They adjust the exhaust fan operation of the furnace only based on the current furnace pressure deviation and trend. In scenarios involving the joint operation of a group of calcium carbide furnaces, the following problems still exist: existing technologies only achieve automatic adjustment of the furnace pressure of a single furnace, failing to treat multiple calcium carbide furnaces as a unified gas source system for coordinated furnace group control, and cannot allocate load according to the operating status of each furnace; furthermore, existing technologies only focus on whether the current furnace pressure is within the set range when performing furnace pressure analysis. Within a certain range, the lack of information on gas consumption trends and planned shutdowns from downstream gas-consuming units such as acetylene plants makes it impossible to adjust furnace pressure in advance based on future gas demand, easily leading to insufficient gas supply. Although existing technologies analyze furnace pressure and link it to fan regulation, they do not comprehensively evaluate multi-source operating parameters such as furnace pressure, current, and alarm records. They lack a furnace health index to quantitatively characterize the operating status of a single calcium carbide furnace, making it impossible to achieve differentiated load distribution in furnace group control. Under long-term changes in operating conditions, there is a lack of adaptive correction and abnormal rollback mechanisms. When prediction deviations increase or individual calcium carbide furnaces malfunction, the system can only rely on manual intervention, making it difficult to guarantee the long-term stable operation of the furnace group and the safety of gas supply.

[0004] To address the aforementioned issues, this invention proposes an intelligent operation control method for calcium carbide furnace pressure. This method involves continuously collecting and short-term forecasts of the calcium carbide furnace group's operation and downstream gas consumption conditions; constructing a multi-indicator weighted furnace health index and calculating the load sharing coefficient accordingly; generating target pressure values ​​and coordinated adjustment commands for each calcium carbide furnace; and dynamically adjusting the gas consumption prediction model and load allocation strategy using adaptive correction and anomaly fallback mechanisms. This achieves forward-looking control of calcium carbide furnace pressure, differentiated load allocation across the furnace group, and long-term safe and stable operation. Summary of the Invention

[0005] In view of the aforementioned existing problems, the present invention is proposed.

[0006] This invention provides an intelligent operation control method for calcium carbide furnace pressure, aiming to solve the problems in the prior art where furnace pressure relies solely on passive adjustment of a single furnace, the load distribution of the furnace group is rigid, and there is a lack of downstream gas consumption prediction and adaptive control for abnormal operating conditions.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0008] A method for intelligent operation control of calcium carbide furnace pressure includes:

[0009] Step S1: Generate basic data on furnace group operation and downstream gas consumption by continuously collecting data on the operation of the calcium carbide furnace group.

[0010] Step S2: Preprocess the basic data to generate downstream gas consumption data series, and use trend analysis and short-term load forecasting to calculate and generate gas demand forecast results.

[0011] Step S3: Based on the gas demand forecast results, analyze the furnace operation parameters and alarm records, calculate the furnace health index of each calcium carbide furnace using multi-index weighting, and perform normalization processing to generate furnace health data.

[0012] Step S4: Based on the gas demand forecast results and furnace health data, decompose the target total furnace pressure and total furnace gas volume, calculate the load sharing coefficient of each calcium carbide furnace, and generate the target furnace pressure value and corresponding adjustment parameters for each calcium carbide furnace.

[0013] Step S5: Based on the target furnace pressure and adjustment parameters, generate adjustment commands for each calcium carbide furnace and perform coordinated adjustment of the calcium carbide furnaces in the furnace group;

[0014] Step S6 involves real-time monitoring of the deviation between predicted and actual gas consumption and changes in the furnace health index, and using adaptive correction and abnormal rollback to dynamically adjust the gas consumption prediction model and load sharing coefficient.

[0015] As a preferred embodiment, the specific steps for generating basic data on furnace group operation and downstream gas consumption by continuously collecting data on the operation of the calcium carbide furnace group are as follows:

[0016] By adopting a layered and point-based continuous data acquisition method for the calcium carbide furnace group, a furnace pressure transmitter, current and voltage detection device, furnace gas flow meter, feed rate metering device and furnace top temperature sensor are installed in each calcium carbide furnace. Pipeline pressure and flow sensors are installed in the main furnace gas pipe and each branch. The above sensor signals are sampled in real time at a fixed sampling period through the field acquisition terminal, and each sampled data is appended with furnace number, sampling timestamp and operating condition identifier, and uploaded to the host computer via industrial Ethernet.

[0017] As a preferred embodiment, the specific steps of generating basic data on furnace group operation and downstream gas consumption by continuously collecting data on the operation of the calcium carbide furnace group further include:

[0018] The collected furnace operation parameters and downstream gas consumption parameters are formatted and preliminarily verified in the host computer, obvious erroneous data are removed, and the data is stored in the historical database in chronological order to obtain basic data on furnace group operation and downstream gas consumption for subsequent prediction and evaluation.

[0019] As a preferred embodiment, the specific steps for generating gas demand forecast results by using trend analysis and short-term load forecasting are as follows:

[0020] By filtering pipeline pressure, branch flow, and gas-consuming device operating condition signals related to downstream gas consumption in chronological order on the host computer, missing values ​​are filled in using adjacent interpolation, and obvious erroneous data are corrected using median replacement, a downstream gas consumption data sequence arranged at uniform time intervals is generated. The average value and slope of the downstream gas consumption data sequence are calculated within a sliding window, where the average value is used as the current load level and the slope is used as an indicator of load increase or decrease trend.

[0021] As a preferred embodiment, the calculation of gas demand forecast results using trend analysis and short-term load forecasting further includes:

[0022] Downstream gas consumption data sequences are predicted using trend analysis and short-term load forecasting methods. The prediction results are then corrected by incorporating planned shutdown information from downstream units, generating gas demand forecasts for subsequent coordinated boiler group control. The exponential smoothing algorithm used in the short-term load forecasting is as follows:

[0023] ,

[0024] in Let be the predicted gas consumption at the a-th sampling time. To predict gas consumption, 'a' represents the sampling time. P is the smoothing coefficient, and P is the actual gas consumption. This is the dot product symbol.

[0025] As a preferred embodiment, the specific steps for calculating the furnace health index of each calcium carbide furnace using a multi-index weighted average and then performing normalization processing to generate furnace health data are as follows:

[0026] By selecting furnace pressure, current, voltage, furnace gas flow, electrode consumption, furnace top temperature, and alarm records from the dataset corresponding to each calcium carbide furnace as furnace condition evaluation inputs, and using time window statistics and feature extraction methods on the above input data, the furnace condition evaluation index of each calcium carbide furnace in the current prediction period is calculated. Each evaluation index is then normalized to map data of different dimensions to the interval between zero and one, and weighting coefficients are set according to the importance of furnace pressure stability, current stability, electrode consumption, and alarm status. A weighted summation method is used to calculate the furnace condition health index of each calcium carbide furnace, obtaining an initial furnace condition health result between zero and one. After obtaining the initial furnace condition health result, the initial furnace condition health results of all calcium carbide furnaces are normalized again to generate furnace condition health data that can be used for furnace group collaborative control.

[0027] As a preferred embodiment, the specific steps for decomposing the target total furnace pressure and total furnace gas volume based on the gas demand forecast results and furnace health data, calculating the load sharing coefficient for each calcium carbide furnace, and generating the target furnace pressure value and corresponding adjustment parameters for each calcium carbide furnace are as follows:

[0028] First, the target total gas volume and target total furnace pressure are calculated in the host computer based on the predicted total gas consumption and the downstream allowable furnace pressure range. The furnace health index of each calcium carbide furnace is normalized. Combined with the rated capacity, current load and maintenance status of each calcium carbide furnace, the adjustable capacity coefficient of each calcium carbide furnace is calculated. The target total gas volume and target total furnace pressure are weighted and decomposed according to the adjustable capacity coefficient to calculate the load sharing coefficient of each calcium carbide furnace. Based on this, the target furnace pressure, target fan speed and target flue damper opening value of each calcium carbide furnace are determined.

[0029] As a preferred embodiment, the specific steps of decomposing the target total furnace pressure and total furnace gas volume based on the gas demand forecast results and furnace health data, calculating the load sharing coefficient of each calcium carbide furnace, and generating the target furnace pressure value and corresponding adjustment parameters for each calcium carbide furnace further include:

[0030] Upper and lower limit constraints and rate of change constraints are applied to the target values ​​of furnace pressure and actuators to generate adjustment parameters that meet the conditions for safe operation;

[0031] The algorithm for the load sharing factor is defined as follows:

[0032] T1, let there be n calcium carbide furnaces in the furnace group; the furnace health index of the nth calcium carbide furnace is... ; Let be the load sharing coefficient for the nth calcium carbide furnace;

[0033] T2. To normalize the furnace health index of each calcium carbide furnace, the upper limit of the furnace health index is first calculated. and lower limit value And normalize the health index of each calcium carbide furnace;

[0034] T3, based on normalization, calculates the load sharing factor, where the algorithm formula for the load sharing factor is as follows:

[0035] ,

[0036] in Let be the load sharing factor for the nth calcium carbide furnace. Let n be the load sharing coefficient for the calcium carbide furnace, and n be the total number of calcium carbide furnaces in the furnace group. The furnace health index of the nth calcium carbide furnace. The furnace health index of the calcium carbide furnace. The summation symbol is 'i', where 'i' is the number of the calcium carbide furnace.

[0037] After calculating the load sharing coefficient of the calcium carbide furnace group, a set of load sharing coefficients for the calcium carbide furnace group is obtained. Based on the set of load sharing coefficients, the target furnace gas volume and corresponding furnace pressure target values ​​for each calcium carbide furnace are decomposed. The decomposition formula is as follows:

[0038] ,

[0039] in For the target furnace gas volume, Let be the load sharing factor for the nth calcium carbide furnace. Let n be the load sharing coefficient for the calcium carbide furnace, and n be the total number of calcium carbide furnaces in the furnace group. The dot product symbol. The target is the total gas volume of the furnace.

[0040] As a preferred embodiment, the specific steps for generating adjustment commands for each calcium carbide furnace based on the furnace pressure target value and adjustment parameters, and for performing coordinated adjustment of the calcium carbide furnaces in the furnace group, are as follows:

[0041] Based on the generated target pressure values ​​and adjustment parameters of each calcium carbide furnace, the corresponding exhaust fan speed setting value, flue damper opening setting value, and electrode lifting displacement are calculated in the host computer according to the target pressure deviation, load sharing coefficient, and actuator response characteristics, using a graded limiting and closed-loop correction method, thereby generating adjustment commands for each calcium carbide furnace.

[0042] As a preferred embodiment, the specific steps for real-time monitoring of the deviation between predicted and actual gas consumption and changes in the furnace health index, and for dynamically adjusting the gas consumption prediction model and load allocation coefficient using adaptive correction and anomaly rollback, are as follows:

[0043] The system reads the actual gas consumption, furnace pressure feedback value, and current furnace health index of each calcium carbide furnace in real time at a preset sampling period in the host computer. It also calculates the average absolute value of the deviation sequence between the gas demand prediction result and the actual gas consumption within a sliding time window to generate a corrected prediction result. At the same time, it calculates the time series change of the furnace health index of each calcium carbide furnace to generate the corresponding furnace pressure target value and adjustment parameters and sends them down for execution. Based on the preset abnormality criteria, it monitors and judges the duration of furnace pressure exceeding the limit, the fault status of key equipment, and the emergency shutdown signal of downstream gas-consuming devices. When the abnormality criteria are met, it triggers the abnormal backoff control strategy and switches to the safe voltage control mode.

[0044] Beneficial effects

[0045] 1. By preprocessing and exponentially smoothing short-term load forecasting of downstream gas consumption data sequences, historical gas consumption and current gas consumption are weighted and superimposed to generate gas demand forecast results for future control cycles. This improves the furnace pressure control target from a passive adjustment based solely on current furnace pressure and current gas consumption to a forward-looking adjustment based on gas consumption trends, significantly reducing frequent and large fluctuations in furnace pressure targets and solving the problems of unstable forecasting, lagging adjustment, gas supply fluctuations, and high energy consumption caused by existing technologies.

[0046] 2. By employing a multi-index weighted and normalized method based on the furnace pressure, current, electrode consumption, and alarm records of each calcium carbide furnace, a furnace health index is calculated. Furthermore, a load-sharing algorithm decomposes the target total furnace gas volume and target total furnace pressure into target furnace pressure values ​​and adjustment parameters for each calcium carbide furnace. This allows furnaces with better conditions to bear more load, while furnaces with poorer conditions automatically reduce their load or exit the current adjustment. This solves the problem in existing technologies where load is allocated according to a fixed ratio, failing to dynamically adjust the load share based on actual furnace conditions. Consequently, the safety and overall energy efficiency of the furnace group's coordinated control are improved.

[0047] 3. By using a sliding time window statistical method and threshold judgment to analyze the deviation between predicted and actual gas consumption and changes in the furnace health index, the parameters of the gas consumption prediction model and the load sharing coefficient are dynamically corrected. When the furnace pressure exceeds the limit for an extended period or a critical equipment failure is detected, the system automatically switches to a safe furnace pressure control mode and limits the maximum adjustment range of each actuator, generating a furnace pressure target value that meets the safe range. This solves the problems of control parameters gradually becoming mismatched with changes in operating conditions and the lack of a unified abnormal backoff strategy in the existing technology, thus ensuring the stability and safety of the calcium carbide furnace group under long-term operation and sudden operating conditions. Attached Figure Description

[0048] Figure 1 This is a flowchart of the present invention.

[0049] Figure 2This is a comparison chart of the technical effects of the present invention, in which the black bars represent the present invention and the gray bars represent the prior art. Detailed Implementation

[0050] To make the technical means, creative features, and achieved objectives and effects of this invention easier to understand, the invention is further described below with reference to specific embodiments. However, the following embodiments are merely preferred embodiments of this invention and not all of them. Other embodiments obtained by those skilled in the art based on the embodiments described herein without creative effort are all within the protection scope of this invention. Unless otherwise specified, the experimental methods in the following embodiments are conventional methods, and the materials and reagents used in the following embodiments are commercially available unless otherwise specified.

[0051] Example 1, combined with Figure 1 The flowchart shown below illustrates an intelligent operation control method for calcium carbide furnace pressure. The specific implementation steps are as follows:

[0052] Step S1: Generate basic data on furnace group operation and downstream gas consumption by continuously collecting data on the operation of the calcium carbide furnace group.

[0053] Step S1 is used to build the basic data environment required for subsequent intelligent control. By continuously collecting the furnace operation parameters and downstream gas consumption parameters of the calcium carbide furnace group, complete, continuous and traceable basic data of furnace group operation and downstream gas consumption are formed, providing data support for subsequent gas demand prediction, furnace health assessment and furnace group collaborative control.

[0054] Specifically, by adopting a layered and point-based continuous data acquisition method for the calcium carbide furnace group, each calcium carbide furnace is equipped with a furnace pressure transmitter, current and voltage detection device, furnace gas flow meter, feed rate metering device, and furnace top temperature sensor. Pipeline pressure and flow sensors are installed in the main furnace gas pipe and each branch. The above sensor signals are sampled in real time at a fixed sampling period through the field acquisition terminal. Each sampled data is appended with the furnace number, sampling timestamp, and operating condition identifier, and uploaded to the host computer via industrial Ethernet. In the host computer, the collected furnace operating parameters and downstream gas consumption parameters are formatted and preliminarily verified, obvious erroneous data are removed, and the data is stored in the historical database in chronological order to obtain the basic data of furnace group operation and downstream gas consumption for subsequent prediction and evaluation.

[0055] The obviously erroneous data refers to the sampling data in which the sensor self-test status is faulty or communication is interrupted; the sampling data in which the values ​​of furnace pressure, current, furnace gas flow rate, etc. exceed the preset physical threshold; and the sudden change data in which the change amplitude exceeds the preset change threshold in adjacent sampling periods.

[0056] Step S2: Preprocess the basic data to generate downstream gas consumption data series, and use trend analysis and short-term load forecasting to calculate and generate gas demand forecast results.

[0057] Step S2 is used to make short-term trend predictions for downstream gas consumption. By preprocessing and load prediction of downstream gas consumption data series, gas demand prediction results for a period of time in the future are generated, providing a forward-looking target basis for subsequent furnace health assessment and furnace group load allocation.

[0058] Specifically, based on the obtained basic data on boiler group operation and downstream gas consumption, gas demand forecasting results are generated by preprocessing the data. Specifically, the system filters pipeline pressure, branch flow, and gas-consuming device operating condition signals related to downstream gas consumption in chronological order on the host computer. Missing values ​​are filled using adjacent interpolation, and obviously erroneous data is corrected using median replacement. This generates a downstream gas consumption data sequence arranged at the same time interval. The system then calculates the average value and slope of the downstream gas consumption data sequence within a sliding window. The average value is used as the current load level, and the slope is used as an indicator of load increase or decrease trends. For example, when the slope is greater than... When a first threshold is preset, the downstream gas load is determined to be on an upward trend; when the slope is less than a second preset threshold, the downstream gas load is determined to be on a downward trend; when the slope is between the first and second thresholds, the downstream gas load is determined to be basically stable. Based on this, trend analysis and short-term load forecasting methods are used to predict the downstream gas consumption data sequence, and the prediction results are corrected by combining the planned shutdown information of downstream units, generating gas demand forecast results for subsequent furnace group coordinated control. This prediction method solves the problem of lagging furnace pressure regulation caused by passive adjustment based solely on current gas consumption. The exponential smoothing algorithm formula used in the short-term load forecasting is as follows:

[0059] ,

[0060] in Let be the predicted gas consumption at the a-th sampling time. To predict gas consumption, 'a' represents the sampling time. P is the smoothing coefficient, and P is the actual gas consumption. The dot product symbol;

[0061] By constructing this algorithm to weight and superimpose historical gas consumption and current gas consumption, the problem of traditional prediction methods that only use instantaneous values ​​or simple arithmetic averages being too sensitive to abnormal fluctuations and causing instability in furnace pressure regulation targets is solved. Specifically, when predicting the target gas consumption within a short-term time window in the future, the current smoothed value can be used as the prediction result, or multiple time steps in the future can be extrapolated as needed, and the average of the extrapolated results can be used as the gas demand prediction for that time window.

[0062] The aforementioned linear extrapolation refers to a prediction method that calculates the gas consumption at each future time point based on the slope and intercept of a fitted straight line showing the change in gas consumption over time within the current time window.

[0063] The planned shutdown information refers to the shutdown or load reduction arrangements of downstream gas-consuming units in the production plan, including the start time, end time and expected gas consumption changes during the corresponding time period.

[0064] The aforementioned trend analysis and short-term load forecasting method refers to calculating the slope of gas consumption change within a preset sliding time window for downstream gas consumption data series to determine the trend of load increase, decrease, or basic stability, and then using forecasting algorithms such as exponential smoothing and linear extrapolation to calculate the gas consumption within a preset forecast period in the future, thereby generating the corresponding gas demand forecasting results.

[0065] Step S3: Based on the gas demand forecast results, analyze the furnace operation parameters and alarm records, calculate the furnace health index of each calcium carbide furnace using multi-index weighting, and perform normalization processing to generate furnace health data.

[0066] Step S3 is used to quantitatively evaluate the operating status of each calcium carbide furnace. By statistically analyzing the furnace operating parameters and performing multi-index weighted calculations, comparable furnace health data is generated, providing a unified and clear basis for furnace condition evaluation for subsequent load sharing and furnace group coordinated adjustment.

[0067] Specifically, based on the gas demand forecast, furnace pressure, current, voltage, gas flow rate, electrode consumption, furnace top temperature, and alarm records are selected from the dataset corresponding to each calcium carbide furnace as furnace condition evaluation inputs. By employing time window statistics and feature extraction methods on the above input data, furnace condition evaluation indicators for each calcium carbide furnace within the current forecast period are calculated. The time window statistics include: obtaining a furnace pressure stability index by calculating the furnace pressure standard deviation and the number of limit exceedances within a preset time window; obtaining a current stability index by calculating the current fluctuation range and the number of sudden changes; obtaining an electrode consumption index by statistically analyzing the electrode consumption per unit time and the number of electrode drop events; and obtaining an alarm index by statistically analyzing the number of overpressure, underpressure, and related fault alarms of the calcium carbide furnace. Subsequently, each evaluation indicator is normalized to map data of different dimensions to a uniform range of zero to one, and weighting coefficients are set according to the importance of furnace pressure stability, current stability, electrode consumption, and alarm status. For example, the weighting coefficients for furnace pressure stability, current stability, electrode consumption, and alarm status are set to 0.4, 0.3, and 0.2, respectively. The weights are set to 0.1, and the sum of all weighting coefficients is 1. The furnace health index of each calcium carbide furnace is calculated by weighted summation, resulting in an initial furnace health index between zero and one. After obtaining the initial furnace health index, the initial furnace health index of all calcium carbide furnaces is normalized again, so that the furnace health index of calcium carbide furnaces in good condition is close to one and the furnace health index of calcium carbide furnaces in poor condition is close to zero, thereby generating furnace health data that can be used for coordinated control of furnace groups. By calculating the furnace health index, the problem that existing technologies cannot quantitatively compare the operating states of different calcium carbide furnaces and cannot provide a unified evaluation basis for subsequent load allocation is solved.

[0068] The excellent furnace condition refers to the operating state in which the furnace pressure, current, electrode consumption and alarm frequency are all within their respective preset allowable ranges within a preset time window, and meet the corresponding stability index requirements.

[0069] The aforementioned poor furnace condition refers to at least one of the following exceeding the preset allowable range within a preset time window: furnace pressure, current, electrode consumption, or number of alarms.

[0070] Step S4: Based on the gas demand forecast results and furnace health data, decompose the target total furnace pressure and total furnace gas volume, calculate the load sharing coefficient of each calcium carbide furnace, and generate the target furnace pressure value and corresponding adjustment parameters for each calcium carbide furnace.

[0071] Step S4 is used to decompose the target total furnace pressure and total furnace gas volume into the target furnace pressure value and adjustment parameters of each calcium carbide furnace based on the determined gas demand and furnace health status, so as to realize the differentiated load distribution and coordinated control of the calcium carbide furnace group.

[0072] Specifically, based on the gas demand forecast obtained in step S2 and the furnace health data obtained in step S3, the target total gas volume and target total furnace pressure are first calculated in the host computer according to the predicted total gas consumption and the downstream allowable furnace pressure range. The furnace health index of each calcium carbide furnace is then normalized. Combined with the rated capacity, current load, and maintenance status of each calcium carbide furnace, the adjustability coefficient of each calcium carbide furnace is calculated. When the furnace health index of a certain calcium carbide furnace is lower than a preset threshold, the adjustability coefficient of that calcium carbide furnace is limited to the lower limit. Then, the target total gas volume and target total furnace pressure are weighted and decomposed according to the adjustability coefficient to calculate the load sharing coefficient of each calcium carbide furnace. Based on this, the target furnace pressure value, target fan speed value, and target flue damper opening value corresponding to each calcium carbide furnace are determined. Upper and lower limit constraints and rate of change constraints are applied to the target furnace pressure value and the target actuator value to generate adjustment parameters that meet the safe operation conditions.

[0073] The algorithm for the load sharing factor is defined as follows:

[0074] T1, let there be n calcium carbide furnaces in the furnace group; the furnace health index of the nth calcium carbide furnace is... ; Let be the load sharing coefficient for the nth calcium carbide furnace;

[0075] T2. To normalize the furnace health index of each calcium carbide furnace, the upper limit of the furnace health index is first calculated. and lower limit value The system normalizes the health index of each calcium carbide furnace. When the health index of a calcium carbide furnace is lower than the preset lower limit, in order to ensure the safety and stability of the system, the normalization result can be forcibly limited to 0, so that the calcium carbide furnace does not participate in the furnace pressure regulation in this load distribution. When the health index of a calcium carbide furnace is higher than the preset upper limit, in order to avoid the excessive weight of a single calcium carbide furnace affecting the overall distribution result, the normalization result can be forcibly limited to 1.

[0076] T3, based on normalization, calculates the load sharing coefficient, thereby enabling the reasonable allocation of the target total furnace pressure and target total furnace gas volume according to the health status of each calcium carbide furnace. This allows calcium carbide furnaces in good condition to bear more load, while those in poor condition bear less load or do not participate in load regulation, thus achieving efficient and stable operation of the furnace group. The algorithm formula for the load sharing coefficient is as follows:

[0077] ,

[0078] in Let be the load sharing factor for the nth calcium carbide furnace. Let n be the load sharing coefficient for the calcium carbide furnace, and n be the total number of calcium carbide furnaces in the furnace group. The furnace health index of the nth calcium carbide furnace. The furnace health index of the calcium carbide furnace. The summation symbol is 'i', where 'i' is the number of the calcium carbide furnace.

[0079] After calculating the load sharing coefficient of the calcium carbide furnace group, a set of load sharing coefficients for the calcium carbide furnace group is obtained. Based on the set of load sharing coefficients, the target furnace gas volume and corresponding furnace pressure target values ​​for each calcium carbide furnace are decomposed. The decomposition formula is as follows:

[0080] ,

[0081] in For the target furnace gas volume, Let be the load sharing factor for the nth calcium carbide furnace. Let n be the load sharing coefficient for the calcium carbide furnace, and n be the total number of calcium carbide furnaces in the furnace group. The dot product symbol. The target total boiler gas volume;

[0082] The load allocation algorithm based on the furnace health index solves the problems in the existing technology of being unable to dynamically allocate the load according to the actual operating status of the calcium carbide furnace, resulting in long-term overload operation of individual calcium carbide furnaces and low overall adjustment efficiency of the furnace group. By normalizing the health index and calculating the load allocation coefficient, the algorithm addresses these issues.

[0083] Step S5: Based on the target furnace pressure and adjustment parameters, generate adjustment commands for each calcium carbide furnace and perform coordinated adjustment of the calcium carbide furnaces in the furnace group;

[0084] Step S5 is used to convert the target values ​​and adjustment parameters of each calcium carbide furnace determined in step S4 into specific execution instructions, and to perform coordinated closed-loop adjustment of multiple calcium carbide furnaces so that the actual operating state of the total furnace pressure and total furnace gas volume of the furnace group approaches the target range.

[0085] Specifically, based on the target pressure values ​​and adjustment parameters of each calcium carbide furnace generated in step S4, the corresponding exhaust fan speed setpoint, flue damper opening setpoint, and electrode lifting displacement are calculated in the host computer according to the furnace pressure target deviation, load sharing coefficient, and actuator response characteristics, using a graded limiting and closed-loop correction method. This generates adjustment commands for each calcium carbide furnace. The adjustment commands are then sent to the fan frequency converter, damper actuator, and electrode lifting device through the industrial control network. Furnace pressure feedback values ​​are collected in real time at a preset sampling period. The furnace pressure feedback values ​​are compared with the furnace pressure target values. Based on the comparison results, the adjustment commands are incrementally corrected and collaboratively optimized so that the total furnace pressure and total furnace gas volume of the furnace group smoothly converge to the target range while meeting the downstream gas demand.

[0086] The aforementioned graded limiting and closed-loop correction method refers to first applying segmented upper limit constraints to the adjustment amplitude and adjustment rate according to a pre-set graded limiting rule when generating the actuator adjustment command, and then collecting the furnace pressure feedback value at a preset sampling period after the adjustment is executed. The furnace pressure feedback value is compared with the furnace pressure target value, and the subsequent adjustment command is incrementally corrected based on the comparison result, thereby forming a closed-loop control process that limits the adjustment amplitude and continuously corrects based on feedback error.

[0087] Step S6: Real-time monitoring of the deviation between predicted gas consumption and actual gas consumption and changes in the furnace health index; dynamic adjustment of the gas consumption prediction model and load sharing coefficient using adaptive correction and abnormal rollback.

[0088] In step S6, the gas consumption prediction deviation and furnace health index changes are evaluated in real time during system operation, the gas consumption prediction model and load sharing coefficient are automatically corrected, and the system switches to safe furnace pressure control mode when abnormal conditions occur, thereby ensuring the long-term stability and safe operation of the calcium carbide furnace group.

[0089] Specifically, based on the data acquisition and control closed loop established in steps S1 to S5, the actual gas consumption, furnace pressure feedback value, and current furnace health index of each calcium carbide furnace are read in real time in the host computer at a preset sampling period. The average absolute value of the deviation sequence between the gas demand prediction result and the actual gas consumption calculation is calculated within a sliding time window. When the average deviation exceeds the prediction deviation threshold for multiple consecutive sampling periods, the parameters of the gas consumption prediction model are automatically adjusted, such as adjusting the slope weight, smoothing coefficient, and time window length in linear extrapolation or exponential smoothing algorithms, thereby correcting subsequent prediction results. Simultaneously, by calculating the time series change of the furnace health index of each calcium carbide furnace, when the health index of a certain calcium carbide furnace decreases by more than a preset health decay threshold in a short period, the load sharing coefficient of that calcium carbide furnace is automatically reduced, and the load sharing coefficients of other calcium carbide furnaces are increased proportionally. The load sharing coefficient is used to recalculate the corresponding furnace pressure target value and adjustment parameters and issue them for execution. Furthermore, based on the pre-set abnormality criteria, the duration of furnace pressure exceeding the limit, the fault status of key equipment, and the emergency shutdown signal of downstream gas-consuming devices are monitored and judged. When any abnormal condition occurs, such as the furnace pressure exceeding the limit for an extended period, the failure of a key fan or pressure sensor, or the emergency shutdown of a downstream gas-consuming device, the host computer immediately freezes the current collaborative adjustment strategy, switches to the safe furnace pressure control mode, adjusts the total furnace pressure target value to the safe range, limits the maximum adjustment amplitude and change rate of each actuator, and limits or shuts down some downstream gas-consuming branches according to a preset sequence. Through the above-mentioned adaptive correction and abnormal rollback, the present invention can automatically correct control parameters and quickly enter a safe state under long-term operation and sudden operating conditions, avoiding the accumulation of prediction deviations and the amplification of abnormal operating conditions.

[0090] Combination Figure 2 The figure shows a comparison of the technical effects of an intelligent operation control method for calcium carbide furnace pressure. The black bars represent the technical effects of the present invention, while the gray bars represent the technical effects of existing technologies. Figure 2 It can be seen that the technical effect of the present invention is superior to that of the prior art.

[0091] Example 2 is an intelligent operation control method for calcium carbide furnace pressure based on Example 1. This example is used in conjunction with the facility in Example 1. The specific solution is as follows:

[0092] In step S1, the sampling period for the furnace group operating parameters and downstream gas consumption parameters is set to 8 seconds to collect the calcium carbide furnace operating data;

[0093] In step S2, the sliding time window length for short-term load forecasting is set to 12 minutes, the smoothing coefficient α of the exponential smoothing algorithm is set to 0.2, and the gas demand forecasting result is calculated based on the gas consumption data within the sliding time window.

[0094] In step S3, the weighting coefficients corresponding to furnace pressure stability, current stability, electrode consumption and alarm status are set to 0.4, 0.3, 0.2 and 0.1 respectively, and the lower limit threshold of the furnace health index is set to 0.3. When the furnace health index of a certain calcium carbide furnace is lower than 0.3, the normalized health data of the calcium carbide furnace is set to 0 through normalization processing.

[0095] In step S4, the load sharing coefficient is calculated based on the normalized health data, and the target total furnace gas volume is allocated to the four calcium carbide furnaces according to the load sharing coefficient. When the normalized health data of a certain calcium carbide furnace is 0, the load corresponding to this furnace pressure adjustment is not allocated to that calcium carbide furnace.

[0096] In step S5, adjustment commands for the four calcium carbide furnaces are issued in a grouped and coordinated manner. The calcium carbide furnaces with non-zero furnace health data are grouped together, and the exhaust fan speed setting value and flue damper opening setting value are sent to the calcium carbide furnaces in this group first. The calcium carbide furnaces with zero furnace health data are kept in their original operating settings. The fan speed change rate and the single adjustment range of the damper opening are set to not exceed 5% and 10% of the rated value, respectively.

[0097] In step S6, by setting the gas consumption prediction deviation threshold to 3%, when the absolute value of the gas consumption prediction deviation is greater than 3% in three consecutive prediction cycles, an adaptive correction of the smoothing coefficient and time window length of the exponential smoothing algorithm is triggered; when the furnace pressure of any calcium carbide furnace reaches 90% of the preset safety upper limit or is lower than 90% of the preset safety lower limit, an abnormal rollback control is triggered, and the furnace pressure control mode is switched to safe furnace pressure control mode.

[0098] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for intelligent operation control of calcium carbide furnace pressure, characterized in that: Step S1: Generate basic data on furnace group operation and downstream gas consumption by continuously collecting data on the operation of the calcium carbide furnace group. Step S2: Preprocess the basic data to generate downstream gas consumption data series, and use trend analysis and short-term load forecasting to calculate and generate gas demand forecast results. Step S3: Based on the gas demand forecast results, analyze the furnace operation parameters and alarm records, calculate the furnace health index of each calcium carbide furnace using multi-index weighting, and perform normalization processing to generate furnace health data. Step S4: Based on the gas demand forecast results and furnace health data, decompose the target total furnace pressure and total furnace gas volume, calculate the load sharing coefficient of each calcium carbide furnace, and generate the target furnace pressure value and corresponding adjustment parameters for each calcium carbide furnace. Step S5: Based on the target furnace pressure and adjustment parameters, generate adjustment commands for each calcium carbide furnace and perform coordinated adjustment of the calcium carbide furnaces in the furnace group; Step S6 involves real-time monitoring of the deviation between predicted and actual gas consumption and changes in the furnace health index, and using adaptive correction and abnormal rollback to dynamically adjust the gas consumption prediction model and load sharing coefficient.

2. The intelligent operation control method for calcium carbide furnace pressure according to claim 1, characterized in that: The specific steps for generating basic data on furnace operation and downstream gas consumption by continuously collecting data on the operation of the calcium carbide furnace group are as follows: By adopting a layered and point-based continuous data acquisition method for the calcium carbide furnace group, a furnace pressure transmitter, current and voltage detection device, furnace gas flow meter, feed rate metering device and furnace top temperature sensor are installed in each calcium carbide furnace. Pipeline pressure and flow sensors are installed in the main furnace gas pipe and each branch. The above sensor signals are sampled in real time at a fixed sampling period through the field acquisition terminal, and each sampled data is appended with furnace number, sampling timestamp and operating condition identifier, and uploaded to the host computer via industrial Ethernet.

3. The intelligent operation control method for calcium carbide furnace pressure according to claim 2, characterized in that: The specific steps for generating basic data on furnace operation and downstream gas consumption by continuously collecting data on the operation of the calcium carbide furnace group also include: The collected furnace operation parameters and downstream gas consumption parameters are formatted and preliminarily verified in the host computer, obvious erroneous data are removed, and the data is stored in the historical database in chronological order to obtain basic data on furnace group operation and downstream gas consumption for subsequent prediction and evaluation.

4. The intelligent operation control method for calcium carbide furnace pressure according to claim 1, characterized in that: The specific steps for generating gas demand forecast results using trend analysis and short-term load forecasting are as follows: By filtering pipeline pressure, branch flow, and gas-consuming device operating condition signals related to downstream gas consumption in chronological order on the host computer, missing values ​​are filled in using adjacent interpolation, and obvious erroneous data are corrected using median replacement, a downstream gas consumption data sequence arranged at uniform time intervals is generated. The average value and slope of the downstream gas consumption data sequence are calculated within a sliding window, where the average value is used as the current load level and the slope is used as an indicator of load increase or decrease trend.

5. The intelligent operation control method for calcium carbide furnace pressure according to claim 4, characterized in that: The method of using trend analysis and short-term load forecasting to generate gas demand forecast results also includes: Downstream gas consumption data sequences are predicted using trend analysis and short-term load forecasting methods. The prediction results are then corrected by incorporating planned shutdown information from downstream units, generating gas demand forecasts for subsequent coordinated boiler group control. The exponential smoothing algorithm used in the short-term load forecasting is as follows: ， in Let be the predicted gas consumption at the a-th sampling time. To predict gas consumption, 'a' represents the sampling time. P is the smoothing coefficient, and P is the actual gas consumption. This is the dot product symbol.

6. The intelligent operation control method for calcium carbide furnace pressure according to claim 1, characterized in that: The specific steps for calculating the furnace health index of each calcium carbide furnace using a multi-index weighted average and then normalizing it to generate furnace health data are as follows: By selecting furnace pressure, current, voltage, furnace gas flow rate, electrode consumption, furnace top temperature, and alarm records from the dataset corresponding to each calcium carbide furnace as furnace condition evaluation inputs, and using time window statistics and feature extraction methods on the above input data, the furnace condition evaluation index of each calcium carbide furnace in the current prediction period is calculated. Each evaluation index is then normalized to map data of different dimensions to the interval between zero and one, and weighting coefficients are set according to the importance of furnace pressure stability, current stability, electrode consumption, and alarm status. Finally, a weighted summation method is used to calculate the furnace condition health index of each calcium carbide furnace, obtaining an initial furnace condition health result between zero and one. After obtaining the initial results of furnace health, the initial results of furnace health for all calcium carbide furnaces are normalized again to generate furnace health data that can be used for collaborative control of the furnace group.

7. The intelligent operation control method for calcium carbide furnace pressure according to claim 1, characterized in that: The specific steps for decomposing the target total furnace pressure and total furnace gas volume based on the gas demand forecast results and furnace health data, calculating the load sharing coefficient for each calcium carbide furnace, and generating the target furnace pressure value and corresponding adjustment parameters for each calcium carbide furnace are as follows: First, the target total gas volume and target total furnace pressure are calculated in the host computer based on the predicted total gas consumption and the downstream allowable furnace pressure range. The furnace health index of each calcium carbide furnace is normalized. Combined with the rated capacity, current load and maintenance status of each calcium carbide furnace, the adjustable capacity coefficient of each calcium carbide furnace is calculated. The target total gas volume and target total furnace pressure are weighted and decomposed according to the adjustable capacity coefficient to calculate the load sharing coefficient of each calcium carbide furnace. Based on this, the target furnace pressure, target fan speed and target flue damper opening value of each calcium carbide furnace are determined.

8. The intelligent operation control method for calcium carbide furnace pressure according to claim 7, characterized in that: The specific steps for decomposing the target total furnace pressure and total furnace gas volume based on the gas demand forecast results and furnace health data, calculating the load sharing coefficient for each calcium carbide furnace, and generating the target furnace pressure value and corresponding adjustment parameters for each calcium carbide furnace also include: Upper and lower limit constraints and rate of change constraints are applied to the target values ​​of furnace pressure and actuators to generate adjustment parameters that meet the conditions for safe operation; The algorithm for the load sharing factor is defined as follows: T1, let there be n calcium carbide furnaces in the furnace group; the furnace health index of the nth calcium carbide furnace is... ; Let be the load sharing coefficient for the nth calcium carbide furnace; T2. To normalize the furnace health index of each calcium carbide furnace, the upper limit of the furnace health index is first calculated. and lower limit value And normalize the health index of each calcium carbide furnace; T3, based on normalization, calculates the load sharing factor, where the algorithm formula for the load sharing factor is as follows: ， in Let be the load sharing factor for the nth calcium carbide furnace. Let n be the load sharing coefficient for the calcium carbide furnace, and n be the total number of calcium carbide furnaces in the furnace group. The furnace health index of the nth calcium carbide furnace. The furnace health index of the calcium carbide furnace. The summation symbol is 'i', where 'i' is the number of the calcium carbide furnace. After calculating the load sharing coefficient of the calcium carbide furnace group, a set of load sharing coefficients for the calcium carbide furnace group is obtained. Based on the set of load sharing coefficients, the target furnace gas volume and corresponding furnace pressure target values ​​for each calcium carbide furnace are decomposed. The decomposition formula is as follows: ， in For the target furnace gas volume, Let be the load sharing factor for the nth calcium carbide furnace. Let n be the load sharing coefficient for the calcium carbide furnace, and n be the total number of calcium carbide furnaces in the furnace group. The dot product symbol. The target is the total gas volume of the furnace.

9. The intelligent operation control method for calcium carbide furnace pressure according to claim 1, characterized in that: The specific steps for generating adjustment commands for each calcium carbide furnace based on the aforementioned furnace pressure target value and adjustment parameters, and for performing coordinated adjustment of the calcium carbide furnaces in the furnace group, are as follows: Based on the generated target pressure values ​​and adjustment parameters of each calcium carbide furnace, the corresponding exhaust fan speed setting value, flue damper opening setting value, and electrode lifting displacement are calculated in the host computer according to the target pressure deviation, load sharing coefficient, and actuator response characteristics, using a graded limiting and closed-loop correction method, thereby generating adjustment commands for each calcium carbide furnace.

10. The intelligent operation control method for calcium carbide furnace pressure according to claim 1, characterized in that: The specific steps for real-time monitoring of the deviation between predicted and actual gas consumption and changes in the furnace health index, and for dynamically adjusting the gas consumption prediction model and load allocation coefficient using adaptive correction and anomaly rollback, are as follows: The system reads the actual gas consumption, furnace pressure feedback value, and current furnace health index of each calcium carbide furnace in real time at a preset sampling period in the host computer. It also calculates the average absolute value of the deviation sequence between the gas demand prediction result and the actual gas consumption within a sliding time window to generate a corrected prediction result. At the same time, it calculates the time series change of the furnace health index of each calcium carbide furnace to generate the corresponding furnace pressure target value and adjustment parameters and sends them down for execution. Based on the preset abnormality criteria, it monitors and judges the duration of furnace pressure exceeding the limit, the fault status of key equipment, and the emergency shutdown signal of downstream gas-consuming devices. When the abnormality criteria are met, it triggers the abnormal backoff control strategy and switches to the safe voltage control mode.

Images