A method and device for analyzing load and adjusting production in the steel industry

Abstract

The application belongs to the technical field of rolling mill control, and discloses a steel industry load analysis and production regulation method and device, which comprises the following steps: obtaining historical load data of a rolling mill and training a load prediction model based on the historical load data; obtaining target load of a target time period according to a power grid demand instruction and the latest load data and inputting the target load into the load prediction model to obtain a predicted load fluctuation interval of the rolling mill in the target time period, and calculating a power fluctuation curve according to historical operation data when it is judged that the target load is not in the predicted load fluctuation interval; inputting the target load into the power fluctuation curve to obtain target operation parameters; judging whether the target operation parameters are in a corresponding preset regulation interval and whether a loss value of the rolling mill under the target operation parameters is lower than a preset upper limit of loss; and adjusting the rolling mill according to the target operation parameters if both conditions are met. The application can optimize the operation mode of the rolling mill to adapt to the power grid demand under the premise of ensuring efficiency and service life.

Description

Technical Field

[0001] This application relates to the field of rolling mill control technology, and in particular to a method and apparatus for load analysis and production regulation in the steel industry. Background Technology

[0002] In the rapidly developing steel industry, the advancement of large-scale and intelligent equipment has significantly improved production efficiency and quality. Therefore, it is crucial to deeply explore and effectively utilize the adjustable load potential of key production equipment such as rolling mills to achieve optimized control of the safe, stable, and economical operation of the power system. Consequently, how to conduct in-depth evaluation and analysis of the accuracy and feasibility of rolling mill load regulation, and thus optimize the rolling mill's operating mode to adapt to the power grid load demand while ensuring equipment efficiency and service life, has become an urgent problem to be solved. Summary of the Invention

[0003] This application provides a method and apparatus for load analysis and production regulation in the steel industry, which can optimize the operation mode of rolling mills to adapt to the load demand of the power grid while ensuring equipment efficiency and service life.

[0004] In a first aspect, embodiments of this application provide a method for load analysis and production regulation in the steel industry, including:

[0005] Obtain historical load data of the rolling mill;

[0006] Train a load prediction model based on historical load data;

[0007] The target load for the target time period is obtained based on the power grid demand command and the latest load data of the rolling mill.

[0008] The latest load data is input into the trained load prediction model to obtain the predicted load fluctuation range of the rolling mill within the target time period, and to determine whether the target load is within the predicted load fluctuation range.

[0009] If not, calculate the power fluctuation curve based on the historical operating data of the rolling mill;

[0010] Input the target load into the power fluctuation curve to obtain the target operating parameters;

[0011] Determine whether the target operating parameters are within the corresponding preset adjustment range, and whether the loss value of the rolling mill is lower than the preset loss upper limit under the target operating parameters; if so, adjust the rolling mill according to the target operating parameters.

[0012] Furthermore, the method also includes:

[0013] After preprocessing the latest load data of the rolling mill, multiple load characteristic parameters are extracted.

[0014] Each load characteristic parameter is combined into a multi-dimensional feature vector;

[0015] The multidimensional feature vector is input into the trained anomaly detection model to obtain an anomaly detection score;

[0016] If the anomaly detection score is lower than the preset danger threshold, the rolling mill will be stopped.

[0017] Furthermore, load characteristic parameters include the average, variance, skewness, peak, and trough values ​​of the latest load data.

[0018] Furthermore, it also includes:

[0019] After acquiring historical load data, a median filtering algorithm is used to denoise the historical load data;

[0020] Linear interpolation is used to fill in missing values ​​in historical load data.

[0021] Furthermore, the method also includes:

[0022] After sending the target operating parameters to the rolling mill's control system, the response delay of the control system is obtained;

[0023] Determine if the response delay exceeds the preset response time;

[0024] If so, the information link for transmitting the target's operating parameters will be optimized.

[0025] Furthermore, the method also includes:

[0026] After sending the target operating parameters, obtain the response power of the rolling mill;

[0027] Determine whether the difference between the response power and the target load is greater than the preset response error;

[0028] If so, the power fluctuation curve will be recalculated based on historical operating data.

[0029] Furthermore, the method also includes:

[0030] The current adjustment range is determined based on the latest load data and the upper limit of the rolling mill's load.

[0031] After obtaining the grid demand command, determine whether the grid demand command contradicts the current regulation range;

[0032] If there is no contradiction, then calculate the target operating parameters.

[0033] Secondly, embodiments of this application provide a load analysis and production regulation device for the steel industry, comprising:

[0034] The acquisition module is used to acquire historical load data of the rolling mill;

[0035] The training module is used to train the load prediction model based on historical load data;

[0036] The target module is used to obtain the target load for the target time period based on the power grid demand command and the latest load data of the rolling mill;

[0037] The prediction module is used to input the latest load data into the trained load prediction model to obtain the predicted load fluctuation range of the rolling mill within the target time period, and to determine whether the target load is within the predicted load fluctuation range.

[0038] The calculation module is used to calculate the power fluctuation curve based on the historical operating data of the rolling mill;

[0039] The adjustment module is used to input the target load into the power fluctuation curve to obtain the target operating parameters;

[0040] The judgment module is used to determine whether the target operating parameters are within the corresponding preset adjustment range, and whether the loss value of the rolling mill is lower than the preset loss limit under the target operating parameters.

[0041] The execution module is used to adjust the rolling mill according to the target operating parameters.

[0042] Furthermore, the device also includes:

[0043] The feature extraction module is used to extract multiple load feature parameters after preprocessing the latest load data of the rolling mill;

[0044] The vector synthesis module is used to combine various load characteristic parameters into a multi-dimensional feature vector;

[0045] The scoring module is used to input multi-dimensional feature vectors into the trained anomaly detection model to obtain anomaly detection scores;

[0046] The rolling mill control module is used to stop the rolling mill from running when the abnormality detection score is lower than the preset danger threshold.

[0047] Furthermore, the device also includes:

[0048] The delay acquisition module is used to acquire the response delay of the control system after sending the target operating parameters to the control system of the rolling mill, and to determine whether the response delay exceeds the preset response time.

[0049] The optimization module is used to optimize the information link that transmits target operating parameters when the preset response time is exceeded.

[0050] Furthermore, the device also includes:

[0051] The response power acquisition module is used to acquire the response power of the rolling mill after sending the target operating parameters;

[0052] The response error judgment module is used to determine whether the difference between the response power and the target load is greater than the preset response error. If so, it returns to the calculation module.

[0053] Furthermore, the device also includes a preprocessing module, which, after acquiring historical load data, uses a median filtering algorithm to denoise the historical load data and performs linear interpolation to supplement missing values ​​in the historical load data.

[0054] Furthermore, the device also includes:

[0055] The range calculation module is used to obtain the current adjustment range based on the latest load data and the upper limit of the rolling mill load;

[0056] The demand judgment module is used to determine whether the power grid demand command contradicts the current adjustment range after obtaining the power grid demand command; if there is no contradiction, the adjustment module is executed.

[0057] Thirdly, embodiments of this application provide a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it performs the steps of a steel industry load analysis and production regulation method as described in any of the above embodiments.

[0058] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of a steel industry load analysis and production regulation method as described in any of the above embodiments.

[0059] In summary, compared with the prior art, the beneficial effects of the technical solution provided in this application include at least the following:

[0060] This application provides a method for load analysis and production regulation in the steel industry. By continuously monitoring and accurately predicting the load data of rolling mills, it combines the power grid demand instructions with the production activities of rolling mills. This not only provides forward-looking guidance for production scheduling and reduces energy waste and production instability risks caused by load fluctuations, but also optimizes the operation mode of rolling mills to adapt to power grid load demands while ensuring equipment efficiency and service life, through the judgment of preset adjustment ranges and preset loss limits. Attached Figure Description

[0061] Figure 1 A flowchart of a load analysis and production regulation method for the steel industry is provided as an exemplary embodiment of this application.

[0062] Figure 2This application provides a structural diagram of a load analysis and production regulation device for the steel industry, which is an exemplary embodiment of the present application. Detailed Implementation

[0063] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0064] Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0065] Please see Figure 1 This application provides a method for load analysis and production regulation in the steel industry, including:

[0066] Step S11: Obtain historical load data of the rolling mill. Further, before executing step S12, preprocess the acquired historical load data. The preprocessing includes using a median filtering algorithm to denoise the historical load data and performing linear interpolation to fill in missing values ​​in the historical load data.

[0067] Step S12: Train the load prediction model based on historical load data.

[0068] Among them, the SVM model can be selected as the load forecasting model.

[0069] Specifically, this application uses a real-time load monitoring system for the rolling mill to acquire historical load data over a past period. The acquired historical load data undergoes preprocessing, including noise reduction and smoothing, filtering peak and valley noise in the dataset, linear interpolation to fill missing values, adding timestamps and temperature-related features, resulting in preprocessed historical load data. The kernel function and training parameters of the SVM model are configured, an iterative optimization algorithm is selected, and the SVM model is trained using the historical load data. The trained SVM model can then predict the load trend of the rolling mill over a subsequent period by inputting the latest real-time load data.

[0070] Taking a steel rolling mill production line as an example, this rolling mill has a power output of up to 30MW. Historical load power data for this rolling mill was collected over the past two months at a sampling frequency of 5 minutes, yielding over 8000 data samples. This included a small amount of scattered outlier data, such as power instantaneously dropping to zero. To eliminate the impact of noise on the model, a median filtering algorithm was used for denoising and smoothing. Additionally, rolling mill temperature data was added as an auxiliary input dimension. The preprocessed dataset was used to train an SVM model to predict future loads. A radial basis function kernel was used to fit the complex mapping relationship between load and temperature. The kernel parameter gamma = 0.5 and the penalty coefficient C = 15 were set, and iterative optimization was performed in steps of 0.0001 until the model error met the specified requirements. The trained SVM model was then tested: real-time load data from the most recent 50 points was input to predict the rolling mill's power change curve for the next day, yielding the expected load range, with a peak power expected to reach approximately 35MW.

[0071] Step S13: Obtain the target load for the target time period based on the power grid demand command and the latest load data of the rolling mill.

[0072] Step S14: Input the latest load data into the trained load prediction model to obtain the predicted load fluctuation range of the rolling mill within the target time period, and determine whether the target load is within the predicted load fluctuation range.

[0073] Specifically, it retrieves grid demand instructions issued by the grid load dispatching system. When a demand instruction is issued, it parses the instruction content and extracts the target load that needs to be increased or decreased within the target time period.

[0074] Simultaneously, the SVM model is used to predict the load fluctuation range of the rolling mill within the target time period required by the power grid demand command. The target load is compared with the load fluctuation range predicted by the SVM. The degree of matching between the target load and the predicted load fluctuation range is analyzed to determine whether the original production plan of the rolling mill can meet the demand command requirements.

[0075] In practical implementation, for example, if the power grid issues a control command to reduce the load by 2MW, parsing this command reveals that the target control quantity for this demand response is to reduce the rolling mill load from the original 32MW to 30MW within the target time period from 17:00 to 20:00. The predicted load fluctuation range for this period is obtained, showing that under current production conditions, the rolling mill load will fluctuate between 30-34MW within the target time period. Comparing the target load of 30MW with the predicted load fluctuation range, it is determined that the load reduction requirement can be met, therefore the original production plan is maintained. However, if the latest command further requires a reduction of 2MW, maintaining the original plan would not meet the control objective. Therefore, the rolling mill production process needs to be optimized.

[0076] If step S15 is not available, then calculate the power fluctuation curve based on the historical operating data of the rolling mill.

[0077] Step S16: Input the target load into the power fluctuation curve to obtain the target operating parameters.

[0078] Specifically, based on historical operating data of the rolling mill, the variation of its load power with equipment temperature and speed parameters is analyzed, and a power fluctuation curve is fitted. In the demand response scenario, the target load is used as the input constraint for the fitting rate fluctuation curve process. For the input target, the optimal combination of temperature and speed parameters is output to form the target operating parameters that meet the response requirements.

[0079] In the specific implementation process, historical operating data such as the rolling mill temperature (t) and tapping rate (v) are acquired to construct a state-space model and establish a dynamic mapping relationship between temperature, rate, and load power (p): dp / dt = f(p, t, v). An LSTM neural network is used for simulation training to learn the complex nonlinear characteristics of the function f. The training samples include historical operating data for temperature, rate, and power. The loss function is set to mean squared error, and the training step size is 0.001. Through 2000 iterations, an accurate mapping model is obtained that can simulate the power fluctuation curve of the rolling mill. In the response control scenario, the target power is set to a target load of 30MW. An RBF neural network observer is introduced. Based on the state observations before and after the response, the target operating parameters of temperature and rate are output. The temperature (t) is adjusted from 950 degrees Celsius to 920 degrees Celsius, and the tapping rate (v) is reduced from 2.8 m / s to 2.5 m / s. These target operating parameters allow the rolling mill power to smoothly follow the target load of 30MW.

[0080] Step S17: Determine whether the target operating parameters are within the corresponding preset adjustment range, and whether the loss value of the rolling mill is lower than the preset loss upper limit under the target operating parameters.

[0081] Step S18: If yes, adjust the rolling mill according to the target operating parameters.

[0082] If the target operating parameters are within the preset adjustment range and the loss value is lower than the preset loss upper limit, the rolling mill is adjusted. If any condition is not met, the process returns to step S15 to refit the power fluctuation curve and reset the target operating parameters.

[0083] Specifically, the power fluctuation range of the rolling mill before and after adjustment is compared to assess the impact of the target operating parameters on the rolling mill's stability. The maximum and minimum values ​​of temperature and pressure are calculated to determine the preset adjustment range, and it is judged whether the target operating parameters exceed the preset adjustment range. Additionally, the impact of the target operating parameters on rolling mill losses is assessed. If the assessment results show that the target operating parameters meet the rolling mill's power response target, and the system stability and equipment health are within normal limits, then the adjustment strategy is confirmed as effective. Otherwise, the power fluctuation curve is refitted, the target operating parameters are redesigned, and this process is iterated until complete target operating parameters are obtained.

[0084] The above embodiment provides a load analysis and production regulation method for the steel industry. By continuously monitoring and accurately predicting the load data of the rolling mill, it combines the power grid demand command with the production activities of the rolling mill. This not only provides forward-looking guidance for production scheduling and reduces energy waste and production instability risks caused by load fluctuations, but also optimizes the rolling mill operation mode to adapt to the power grid load demand while ensuring equipment efficiency and service life, through the judgment of preset adjustment range and preset loss upper limit.

[0085] In some embodiments, the method further includes:

[0086] Step S21: After preprocessing the latest load data of the rolling mill, extract multiple load characteristic parameters.

[0087] The load characteristic parameters include the average, variance, skewness, peak value, and valley value of the latest load data.

[0088] Step S22: Combine the load characteristic parameters into a multidimensional feature vector.

[0089] Step S23: Input the multidimensional feature vector into the trained anomaly detection model to obtain the anomaly detection score.

[0090] Step S24: If the anomaly detection score is lower than the preset danger threshold, the rolling mill operation is stopped.

[0091] Specifically, the load state is represented by a multi-dimensional feature vector composed of various load characteristic parameters. This multi-dimensional feature vector is input into an anomaly detection model, which can employ the IsolationForest model. Based on an unsupervised learning algorithm, it detects anomalies by isolating outliers. The IsolationForest model determines whether the current multi-dimensional feature vector is abnormal. Once an anomaly is detected, it indicates a problem with the rolling mill load and signs of a fault. An immediate fault warning is issued, and the machine is shut down for inspection and repair to prevent accidents caused by motor overheating or excessive torque. In the specific implementation, the 12 most recent points are extracted from the preprocessed latest load data of the rolling mill using a sliding window to form a feature vector representing the current load state. This feature vector is input into the IsolationForest model for real-time anomaly detection. This model recursively segments the sample feature space and isolates the normal load boundary. The closer a test sample is to the boundary, the lower its score, and it is marked as an abnormal sample. Therefore, in monitoring, a preset danger threshold of 0.3 is set; a model score below 0.3 is considered abnormal.

[0092] The above embodiments enable real-time monitoring of the rolling mill's operating status, quickly identify equipment faults, and take corresponding measures to minimize unexpected downtime and production interruption losses.

[0093] Furthermore, the method also includes:

[0094] After sending the target operating parameters to the rolling mill's control system, the response delay of the control system is obtained; it is then determined whether the response delay exceeds the preset response time. If so, the information link for transmitting the target operating parameters is optimized.

[0095] Specifically, during the load adjustment of the rolling mill, the response delay of the production execution system to control commands is monitored in real time. The response delay is calculated by comparing the timestamp of the adjustment command issuance time with the actual system response time. This is then compared with a preset response time. If the monitored actual response delay exceeds the preset response time, the production execution system's response speed is deemed too slow and cannot meet the real-time requirements of the adjustment strategy. In this case, the information transmission process of the production execution system is optimized. Optimization measures include upgrading the bandwidth of the internal communication network, simplifying transmission links, reducing transmission nodes, and optimizing the command parsing algorithm rate. After process optimization, the load adjustment strategy is re-executed, and the response delay is monitored to determine if the preset response time is met. If the real-time requirements are still not met, process weaknesses are identified and iteratively optimized until the response delay meets the standard.

[0096] In the specific implementation process, pre-defined target operating parameters such as temperature and speed were issued to the rolling mill production line, with a load control target of 30MW. This instruction was sent to the production execution system via industrial Ethernet at a rate of 100ms. However, the actual response latency of the system was monitored to be as high as 300ms, exceeding the preset response time by 150ms and far exceeding the network transmission rate. Analysis indicated that the receiving, parsing, and processing speed of the production system was too slow, failing to meet the real-time adjustment requirements. Optimization measures included upgrading the controller of the execution system, adopting a parallel multi-core GPU architecture to improve the instruction interpretation and processing speed, simplifying the control logic, eliminating the intermediate conversion and parsing link in the system, and directly writing the instruction value into the OPC interface; adjusting the temperature setting loop, using high-speed FPGA pulse control instead of PLC to achieve fast and accurate positioning. After process optimization, the monitoring execution system response latency was reduced to 120ms, meeting the real-time requirements. Ultimately, the optimized system responded to the set speed and temperature values ​​within 30ms according to the adjustment strategy, reading the casting strip speed reduced to 2.5 m / s, the temperature at 920 degrees Celsius, and the rolling mill load power maintained in real time between 29.8-30.2MW.

[0097] Furthermore, the method also includes:

[0098] After sending the target operating parameters, the response power of the rolling mill is obtained; it is determined whether the difference between the response power and the target load is greater than the preset response error; if so, the power fluctuation curve is recalculated based on historical operating data.

[0099] Specifically, data on the rolling mill after implementing the regulation strategy is obtained from the power management platform, including information on temperature, rate, and response power. The changes in average power, fluctuation range, and peak-to-valley difference before and after regulation are compared. If the comparison results show that the deviation between the actual response power and the target load is greater than the preset response error, the equipment status is abnormal, indicating poor environmental matching of the regulation strategy. It is necessary to refit the power fluctuation curve, design target operating parameters, and iterate until a complete regulation strategy is obtained.

[0100] Furthermore, specific analysis can be conducted on the equipment status data of temperature and pressure after the rolling mill implements the target operating parameters, detecting any exceeding of limits and assessing the impact of the adjustment process on equipment health. Business data on rolling mill output can be read to assess the overall impact of adjustments on production. The compatibility between load power adjustment strategies and the actual environment can also be evaluated.

[0101] In the specific implementation process, the average power after adjustment obtained from the power system management platform was 30.5MW, with a fluctuation range of 29MW-32MW. Compared with the target load of 30MW, the control deviation was within the allowable range.

[0102] Further analysis of the temperature data revealed a peak temperature of 978 degrees Celsius, exceeding the equipment's safety limit of 965 degrees Celsius. Simultaneously, a 15% exceedance of the bearing pressure was detected. Therefore, this load adjustment caused some wear and tear on the equipment.

[0103] The above embodiments, by judging the difference between the actual response power of the rolling mill to the target operating parameters and the target load, can provide feedback guidance for the generation of the target operating parameters, making the load adjustment of the rolling mill more accurate.

[0104] Furthermore, the method also includes:

[0105] The current adjustment range is obtained based on the latest load data and the upper limit of the rolling mill load; after obtaining the power grid demand command, it is determined whether the power grid demand command contradicts the current adjustment range; if they do not contradict each other, the target operating parameters are calculated.

[0106] Specifically, based on the latest real-time load data, statistical parameters such as average load, fluctuation range, and peak-valley values ​​are determined. If the load remains high and reaches the equipment's power limit, it is considered a full-load operating state; if the load remains at a mid-to-lower level, it is considered a light-load state. For different operating states, the current load adjustment range is obtained. Upon receiving a grid demand command, it is determined whether the current adjustment range can meet the grid demand command. If it does, the target operating parameters are calculated.

[0107] During implementation, the latest load data collected in real time from the rolling mill shows that its current average load is 31MW, with peak and valley power fluctuating between 25-32MW, indicating full-load operation. Based on historical statistical analysis, the rolling mill's rated power is 31MW, meaning it still has approximately 3MW of downward adjustment space under full load, determining its current adjustment range to be 28MW-31MW. If the grid demand command requires a 2MW reduction, i.e., a target load of 29MW, within the current adjustment range, steps S14-S18 in the above embodiment are executed to obtain the target operating parameters.

[0108] Furthermore, this application can also monitor the execution of power grid demand commands, and summarize the load data and adjustment process of the adjusted rolling mill in a unified manner, and feed them back to the power grid operator and plant management personnel.

[0109] Specifically, after generating the target operating parameters, this application will construct specific scenario application instructions based on the target operating parameters and the corresponding production plan. The scenario application instructions include the target value of the die temperature of the rolling mill, the value of the rotational speed, and parameters for thrust and casting rate. These instructions are sent to the rolling mill via industrial Ethernet at a rate of 100ms, and the system's response dynamics are monitored in real time. Once an adjustment command is detected from the rolling mill for the relevant equipment parameters, the sensor data of the rolling mill is monitored to track whether the parameters have been adjusted according to the requirements of the scenario application instructions. It is verified whether each parameter has been adjusted to the value required by the target operating parameters. If, during monitoring, a mismatch is found between the actual response and the instruction requirements, the scenario application instructions are resent for the relevant components until a complete response is achieved. Upon completion, the constructed scenario application instructions, execution process dynamic information, parameter response data, and a summary report are compiled and sent to the power grid operator and factory management personnel.

[0110] In the specific implementation process, if the target operating parameters are a die temperature of 960℃, a tapping rate of 30 m / s, and a corresponding output of 28,000 tons / hour, then the scenario application commands are: SETTemp = 960℃; SETSpeed ​​= 30 m / s; SETOutput = 28000 t / h. These commands control the temperature measurement module, rate sensor, and output calculation unit. Simultaneously, a real-time monitoring program is initiated to continuously track temperature image data from an industrial camera. Once the detected temperature reaches the set value of 960℃, the command is acknowledged, and the temperature control unit completes the adjustment action. Feedback signals are read from the variable frequency speed control device, and the bearing speed stabilizes at 30 m / s, verifying that the rate has responded to the command requirements. Finally, information is collected from the output database. When the output calculation result is reported as 28,000 tons / hour, it is confirmed that all parameters have been adjusted to optimal settings, and the command execution is complete. The commands and parameter response curves are summarized, an adjustment report is generated, and sent to the power grid and plant management system for centralized feedback of results.

[0111] Please see Figure 2 Another embodiment of this application provides a load analysis and production regulation device for the steel industry, comprising:

[0112] The acquisition module 101 is used to acquire historical load data of the rolling mill.

[0113] Training module 102 is used to train a load prediction model based on historical load data.

[0114] The target module 103 is used to obtain the target load for the target time period based on the power grid demand command and the latest load data of the rolling mill.

[0115] The prediction module 104 is used to input the latest load data into the trained load prediction model to obtain the predicted load fluctuation range of the rolling mill within the target time period, and to determine whether the target load is within the predicted load fluctuation range.

[0116] The calculation module 105 is used to calculate the power fluctuation curve based on the historical operating data of the rolling mill.

[0117] The adjustment module 106 is used to input the target load into the power fluctuation curve to obtain the target operating parameters.

[0118] The judgment module 107 is used to determine whether the target operating parameters are within the corresponding preset adjustment range, and whether the loss value of the rolling mill is lower than the preset loss upper limit under the target operating parameters.

[0119] The execution module 108 is used to adjust the rolling mill according to the target operating parameters.

[0120] Furthermore, when the judgment module 107 determines that the target operating parameter is not within the corresponding preset adjustment range, or that the loss value of the rolling mill under the target operating parameter is higher than the preset loss limit, it returns to the calculation module 105.

[0121] Furthermore, the device also includes:

[0122] The feature extraction module 201 is used to extract multiple load feature parameters after preprocessing the latest load data of the rolling mill.

[0123] The vector synthesis module 202 is used to combine the various load characteristic parameters into a multi-dimensional feature vector.

[0124] The scoring module 203 is used to input multidimensional feature vectors into the trained anomaly detection model to obtain anomaly detection scores.

[0125] The rolling mill control module 204 is used to stop the rolling mill if the abnormality detection score is lower than the preset danger threshold.

[0126] Furthermore, the device also includes:

[0127] The delay acquisition module 301 is used to acquire the response delay of the control system after sending the target operating parameters to the control system of the rolling mill, and to determine whether the response delay exceeds the preset response time.

[0128] The optimization module 302 is used to optimize the information link that transmits target operating parameters when the preset response time is exceeded.

[0129] Furthermore, the device also includes:

[0130] The response power acquisition module 401 is used to acquire the response power of the rolling mill after sending the target operating parameters.

[0131] The response error judgment module 402 is used to determine whether the difference between the response power and the target load is greater than the preset response error. If so, it returns to the calculation module 105 to recalculate the power fluctuation curve based on historical operating data.

[0132] Furthermore, the device also includes a preprocessing module 501, which, after acquiring historical load data, uses a median filtering algorithm to denoise the historical load data and performs linear interpolation to supplement missing values ​​in the historical load data.

[0133] Furthermore, the device also includes:

[0134] The range calculation module 601 is used to obtain the current adjustment range based on the latest load data and the upper limit of the rolling mill load.

[0135] The demand judgment module 602 is used to determine whether the power grid demand command contradicts the current adjustment range after obtaining the power grid demand command; if they do not contradict each other, the adjustment module 106 is executed.

[0136] The specific limitations of the load analysis and production regulation device for the steel industry provided in this embodiment can be found in the embodiment of the load analysis and production regulation method for the steel industry described above, and will not be repeated here. Each module in the aforementioned load analysis and production regulation device for the steel industry can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0137] This application provides a computer device that may include a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface communicates with external terminals via a network connection. When the computer program is executed by the processor, it causes the processor to perform the steps of a steel industry load analysis and production regulation method as described in any of the above embodiments.

[0138] The working process, working details, and technical effects of the computer equipment provided in this embodiment can be found in the embodiment of a load analysis and production adjustment method for the steel industry described above, and will not be repeated here.

[0139] This application provides a computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements the steps of a steel industry load analysis and production regulation method as described in any of the above embodiments. The computer-readable storage medium refers to a data storage medium, which may include, but is not limited to, floppy disks, optical disks, hard disks, flash memory, USB flash drives, and / or memory sticks. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The working process, details, and technical effects of the computer-readable storage medium provided in this embodiment can be found in the embodiments of a steel industry load analysis and production regulation method described above, and will not be repeated here.

[0140] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM).

[0141] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0142] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for load analysis and production regulation in the steel industry, characterized in that, include: Obtain historical load data of the rolling mill; The load prediction model is trained based on the historical load data. The target load for the target time period is obtained based on the power grid demand command and the latest load data of the rolling mill. The latest load data is input into the trained load prediction model to obtain the predicted load fluctuation range of the rolling mill within the target time period, and to determine whether the target load is within the predicted load fluctuation range. If not, the power fluctuation curve is calculated based on the historical operating data of the rolling mill. The calculation includes: analyzing the variation of the load power with the parameters of equipment temperature and speed based on the historical operating data of the rolling mill, and fitting the power fluctuation curve. The target load is input into the power fluctuation curve to obtain the target operating parameters, which include the optimal combination of temperature and rate parameters to form target operating parameters that meet the response requirements. Determine whether the target operating parameters are within the corresponding preset adjustment range, and whether the loss value of the rolling mill is lower than the preset loss upper limit under the target operating parameters; if so, adjust the rolling mill according to the target operating parameters.

2. The method for load analysis and production regulation in the steel industry according to claim 1, characterized in that, Also includes: After preprocessing the latest load data of the rolling mill, multiple load characteristic parameters are extracted. The load characteristic parameters are combined into a multi-dimensional feature vector; The multidimensional feature vector is input into the trained anomaly detection model to obtain an anomaly detection score; If the anomaly detection score is lower than a preset danger threshold, the rolling mill operation will be stopped.

3. The method for load analysis and production regulation in the steel industry according to claim 2, characterized in that, The load characteristic parameters include the average, variance, skewness, peak value, and valley value of the latest load data.

4. The method for load analysis and production regulation in the steel industry according to claim 1, characterized in that, Also includes: After acquiring the historical load data, a median filtering algorithm is used to denoise the historical load data; Missing values ​​in the historical load data are supplemented by linear interpolation.

5. The method for load analysis and production regulation in the steel industry according to claim 1, characterized in that, Also includes: After sending the target operating parameters to the control system of the rolling mill, the response delay of the control system is obtained; Determine whether the response delay exceeds the preset response duration; If so, the information link for transmitting the target operating parameters will be optimized.

6. The method for load analysis and production regulation in the steel industry according to claim 5, characterized in that, Also includes: After sending the target operating parameters, the response power of the rolling mill is obtained; Determine whether the difference between the response power and the target load is greater than a preset response error; If so, the power fluctuation curve is recalculated based on the historical operating data.

7. The method for load analysis and production regulation in the steel industry according to claim 1, characterized in that, Also includes: The current adjustment range is obtained based on the latest load data and the upper limit of the rolling mill load; After obtaining the power grid demand command, determine whether the power grid demand command contradicts the current adjustment range; If there is no contradiction, then calculate the target operating parameters.

8. A load analysis and production regulation device for the steel industry, characterized in that, include: The acquisition module is used to acquire historical load data of the rolling mill; The training module is used to train a load prediction model based on the historical load data; The target module is used to obtain the target load for the target time period based on the power grid demand command and the latest load data of the rolling mill; The prediction module is used to input the latest load data into the trained load prediction model to obtain the predicted load fluctuation range of the rolling mill within the target time period, and to determine whether the target load is within the predicted load fluctuation range. The calculation module is used to calculate the power fluctuation curve based on the historical operating data of the rolling mill. The calculation includes: analyzing the variation of the load power with the parameters of equipment temperature and speed based on the historical operating data of the rolling mill, and fitting the power fluctuation curve. The adjustment module is used to input the target load into the power fluctuation curve to obtain the target operating parameters, which include: the optimal combination of temperature parameters and rate parameters to form target operating parameters that meet the response requirements; The judgment module is used to determine whether the target operating parameters are within the corresponding preset adjustment range, and whether the loss value of the rolling mill is lower than the preset loss upper limit under the target operating parameters. An execution module is used to adjust the rolling mill according to the target operating parameters.

9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the steel industry load analysis and production regulation method as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the steel industry load analysis and production regulation method as described in any one of claims 1 to 7.

Images