Intelligent energy management method and system for intelligent factory

Abstract

The invention relates to the technical field of factory management, in particular to an intelligent factory energy management method and system, and the method comprises the steps: monitoring each monitoring data in a coke preparation process, packaging the monitoring data of each sensor corresponding to the production of single-furnace coke, so as to obtain a plurality of monitoring data packets for the coke of each furnace; preprocessing the data of each monitoring data packet; predicting the expected coke yield based on the cleaned monitoring data; and determining whether the monitoring management for the coke is qualified or not based on the predicted deviation value, and correcting the supervision parameter for the coke based on the detected data fluctuation parameter and the yield change characterization value when it is determined that the monitoring management for the coke is abnormal. And monitoring parameters are dynamically adjusted according to data fluctuation and yield change when monitoring management is abnormal, so that the monitoring efficiency of the intelligent factory is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of factory management, and in particular to an intelligent factory energy management method and system. BACKGROUND

[0002] China is a large producer, consumer and exporter of coking products, which are widely used in chemical industry, pharmaceutical industry, refractory materials industry and national defense industry. In recent years, the coking industry has developed rapidly. At present, the state has made the rational use of resources, energy saving and emission reduction, and environmental protection as a basic national policy, and proposed the strategy of transforming the mode of economic growth and sustainable development. The overall goal of economic and energy development is also proposed in the national economic planning. Therefore, China's energy development strategy should adhere to energy saving and consumption reduction in the long term, improve energy utilization rate, accelerate energy structure adjustment, vigorously develop clean energy and actively develop and utilize renewable energy. At the same time, various measures should be taken to promote the implementation of the strategy.

[0003] The overall industry layout of domestic steel and coking is in the adjustment period. With the gradual development of the market, the influence of national policy on the industry is becoming greater and greater. The "Coking Industry Development Planning Outline" has been completed and officially released. The "Planning Outline" proposes that the coking industry will eliminate all backward production capacity, and the production capacity that meets the access standard will reach more than 70%. At the same time, environmental protection upgrading has become the only way for coking enterprises to survive. In early 2018, the Ministry of Environmental Protection officially issued the "Announcement on the Implementation of Special Emission Limits for Air Pollutants by Cities on the Atmospheric Pollution Transmission Channel of Beijing-Tianjin-Hebei", which requires coking enterprises in "26+2" cities to implement special emission limits for sulfur dioxide, nitrogen oxides, particulate matter and volatile organic compounds from October 1, 2019. With the normalization of environmental protection, some cities not on the "26+2" list also voluntarily implement special emission limit standards.

[0004] A coking plant generally consists of a recovery workshop, a tar processing workshop, a benzene processing workshop, a desulfurization workshop and a wastewater treatment workshop. Accurate monitoring of each workshop to achieve stable coke output is the trend of the times.

[0005] Chinese patent publication No. CN115752576B discloses a coke oven exchange system equipment online monitoring method. When the coke oven is normally heated, the temperature at the time of air exchange is taken as the basis to set the fault determination standard, including coal gas weight chain breakage or waste gas weight card, air cover plate not opened or waste gas weight card, waste gas weight card or air cover plate not closed tightly, exchange not exchanged, exchange machine pull rod broken or not exchanged in place, etc. It can be seen from this that the above technical solution has the following problems: it does not consider dynamically adjusting the supervision parameters according to data fluctuations and yield changes when monitoring management is abnormal, which affects the monitoring efficiency of the intelligent factory. SUMMARY

[0006] To this end, the present application provides an intelligent intelligent factory energy management method and system to overcome the problem that the prior art does not consider dynamically adjusting the monitoring parameters according to data fluctuations and yield changes when monitoring and managing abnormalities, affecting the monitoring efficiency of the intelligent factory.

[0007] In one aspect, the present application provides an intelligent intelligent factory energy management method, comprising: monitoring each monitoring data in the coke preparation process, and packaging the monitoring data of each sensor corresponding to the production of single coke to obtain a plurality of monitoring data packets for each coke; preprocessing the data of each monitoring data packet; predicting the expected coke yield based on each monitoring data after cleaning; determining whether the monitoring management of the coke is qualified based on the predicted deviation value, comprising: determining that the monitoring management of the coke is abnormal, and correcting the monitoring parameters for the coke based on the detection data fluctuation parameter and the yield change characteristic value, including adjusting the monitoring frequency of each sensor to the corresponding value, adjusting the cleaning threshold used to filter out abnormal data in the preprocessing process to the corresponding value, adjusting the coking time in the coke refining process to the corresponding value, or issuing corresponding alarm information; Or, determine that the monitoring management of the coke is qualified, and continue to complete the monitoring management of the coke with the current monitoring parameters.

[0008] Further, the process of determining whether the monitoring management of the coke is qualified based on the predicted deviation value comprises: the absolute value of the difference between the expected coke yield and the corresponding actual monitored coke yield is recorded as the predicted deviation value; when the predicted deviation value is less than or equal to the preset predicted deviation value, it is determined that the monitoring management of the coke is qualified, and the monitoring management of the coke is continued with the current monitoring parameters; when the predicted deviation value is greater than the preset predicted deviation value, it is determined that the monitoring management of the coke is abnormal, and the monitoring parameters for the coke are corrected based on the detection data fluctuation parameter.

[0009] Further, the process of determining the detection data fluctuation parameter comprises: obtain a plurality of monitoring data of a single sensor for a single monitoring data packet, solve the maximum and minimum difference value of each monitoring data, calculate the ratio of the maximum and minimum difference value to the average value of each monitoring data obtained, obtain the data fluctuation value of the single sensor for the single monitoring data packet, and calculate the ratio of the data fluctuation value to the preset data fluctuation value corresponding to the single sensor, to obtain the floating ratio value of the single sensor for the single monitoring data packet. An average of the floating ratio values of the sensors of each data package used to determine the expected coke output is calculated to obtain a data fluctuation parameter.

[0010] Further, the process of correcting the regulatory parameters for coke based on the detected data fluctuation parameter comprises: If the detected data fluctuation parameter is less than or equal to a preset detected data fluctuation parameter, the regulatory parameters for coke are corrected based on the output change representation value; If the detected data fluctuation parameter is greater than the preset detected data fluctuation parameter, the monitoring frequency of each sensor is adjusted to a corresponding value based on the detected data fluctuation parameter.

[0011] Further, the monitoring frequency of the high-risk sensor is adjusted to a corresponding value based on the detected data fluctuation parameter, wherein, The sensor with a floating ratio value greater than a preset floating ratio value is determined as a high-risk sensor; The increase range of the monitoring frequency of the high-risk sensor is positively correlated with the detected data fluctuation parameter.

[0012] Further, when the adjustment of the monitoring frequency of each high-risk sensor is completed, the cleaning threshold of the data processing module for screening abnormal data is adjusted to a corresponding value based on the actual coke output of the single-furnace coke, wherein, The increase range of the cleaning threshold is positively correlated with the actual coke output of the single-furnace coke.

[0013] Further, when the adjustment of the cleaning threshold is completed, a prediction deviation value is determined based on the reacquired monitoring data, and whether the monitoring management for coke is qualified is determined based on the re-determined prediction deviation value, comprising: When the re-determined prediction deviation value is less than or equal to a preset prediction deviation value, it is determined that the monitoring management for coke is qualified, and an alarm information for each high-risk sensor anomaly is issued; When the re-determined prediction deviation value is greater than the preset prediction deviation value, it is determined that the monitoring management for coke is abnormal, and the regulatory parameters for coke are corrected based on the output change representation value.

[0014] Further, the process of correcting the regulatory parameters for coke based on the output change representation value comprises: A prediction deviation value time domain curve is drawn based on the prediction deviation values in the acquired historical data, and a slope of the curve at a current time node is determined as an output change representation value; When the output change representation value is less than or equal to a preset output change representation value, the total amount of coal charged for single-furnace coke refining is adjusted to a corresponding value based on the prediction deviation value; When the output change representation value is greater than the preset output change representation value, an alarm information for raw material anomaly is issued; The decrease range of the total amount of the coal into the furnace is positively correlated with the prediction deviation value.

[0015] Further, when the adjustment of the total amount of the coal into the furnace is completed, it is determined whether to correct the generation parameter of the coke based on the change coefficient, including: The prediction deviation value is determined based on the reacquired monitoring data, and a ratio of the prediction deviation value before the total amount of the coal into the furnace is adjusted is solved to obtain the change coefficient; If the change coefficient is less than or equal to a preset change coefficient, it is determined that the current coke generation parameter is continuously used to complete the refining of the coke; If the change coefficient is greater than the preset change coefficient, the coking time in the coke refining process is adjusted to a corresponding value based on the prediction deviation value; The increase range of the coking time of the coke is positively correlated with the prediction deviation value.

[0016] On the other hand, the present application also provides an intelligent factory energy management system using the above method, including: A coal blending module is used to blend each type of coking coal; A coking module is connected with the coal blending module, and is used to refine each type of coking coal blended by high-temperature treatment into coke; A recovery module is connected with the coking module, and is used to extract the raw coal gas generated by the coking module and purify the extracted raw coal gas; A coke quenching module is connected with the coking module, and is used to cool the coke output by the coking module; A monitoring module is connected with the coal blending module, the coking module and the recovery module, and is used to acquire monitoring data of the coal blending module, the coking module and the recovery module; A data processing module is connected with the monitoring module, and is used to preprocess each monitoring data; A prediction module is connected with the data processing module, and is used to predict the expected coke output based on each monitoring data after cleaning; An analysis module is connected with the monitoring module and the prediction module, and is used to determine whether the monitoring management of the coke is qualified based on the prediction deviation value, and correct the monitoring parameter of the coke based on the detection data fluctuation parameter and the yield change representation value when it is determined that the monitoring management of the coke is abnormal, including adjusting the monitoring frequency of the monitoring module to a corresponding value, adjusting the cleaning threshold value of the data processing module to a corresponding value, adjusting the coking time of the coking module to a corresponding value, or issuing a corresponding alarm information; An alarm module is connected with the analysis module, and is used to issue a corresponding alarm information based on the determination result of the analysis module.

[0017] Compared with the prior art, the present application has the beneficial effects that the monitoring data in the coke preparation process is monitored, the monitoring data of each sensor corresponding to the production single coke oven is packaged to obtain a plurality of monitoring data packets for each coke oven; the data of each monitoring data packet is preprocessed; the expected coke production is predicted based on the cleaned monitoring data; it is determined whether the monitoring management for the coke is qualified based on the prediction deviation value, and when it is determined that the monitoring management for the coke is abnormal, the monitoring parameters for the coke are corrected based on the detection data fluctuation parameter and the yield change characteristic value. When the monitoring management is abnormal, the monitoring parameters are dynamically adjusted according to the data fluctuation and the yield change, thereby improving the monitoring efficiency for the intelligent factory.

[0018] Further, the data fluctuation parameter is determined, including determining a data fluctuation value, a preset data fluctuation value, and a fluctuation comparison value. The data fluctuation value reflects the fluctuation degree of the data of a single sensor in a period of time. The preset data fluctuation value represents the normal fluctuation range of the sensor. The fluctuation comparison value measures the deviation degree of the data fluctuation of a single sensor from the normal range. The data fluctuation parameter comprehensively reflects the overall fluctuation of all the sensor data participating in the calculation.

[0019] Further, the prediction deviation value reflects the accuracy of the yield prediction and reflects the difference between the actual yield and the expected yield. Based on the prediction deviation value, it is determined whether the monitoring management for the coke is qualified. The prediction deviation value provides a quantitative basis for production management. When the prediction deviation value is greater than the preset prediction deviation value, the yield deviation is abnormal. In this case, the data fluctuation parameter is obtained, which comprehensively reflects the overall fluctuation of all the sensor data participating in the calculation. The fluctuation of the sensor monitoring data reflects the stability of the production process. When the detection data fluctuation parameter is greater than the preset detection data fluctuation parameter, the fluctuation is large. In this case, the data accuracy is abnormal due to sensor failure or environmental interference. For this case, the monitoring frequency of each sensor is adjusted to the corresponding value based on the detection data fluctuation parameter to increase the amount of data analysis and improve the accuracy of yield prediction, thereby improving the monitoring efficiency for the intelligent factory. Further, the sensor data with a high fluctuation comparison value has a large fluctuation and may be abnormal. The greater the detection data fluctuation parameter, the more serious the problem. Increasing the monitoring frequency of high-risk sensors ensures timely and accurate grasp of the production state. Concentrating resources on key monitoring of high-risk sensors improves the probability of discovering problems. The monitoring frequency is dynamically adjusted according to the fluctuation degree, the amount of analysis data is increased, the amount of accurate data is increased, the accuracy of data prediction is ensured, and the monitoring efficiency for the intelligent factory is improved.

[0020] Further, after adjusting the high-risk sensor monitoring frequency, the cleaning threshold of the data processing module is adjusted to the corresponding value based on the actual single-furnace coke output, and the increase range of the cleaning threshold is positively correlated with the actual single-furnace coke output; the actual single-furnace coke output reflects the production situation. When the output is high, the data fluctuation is larger, and the cleaning threshold is increased to retain more effective data; when the output is high, the production process changes more, the data fluctuation is larger, and the screening threshold is increased to avoid misjudging effective data as abnormal. According to the output, the cleaning threshold is dynamically adjusted to improve the accuracy and effectiveness of data processing, so that the data processing strategy can adapt to different production states and better reflect the actual production situation. The monitoring efficiency for the intelligent factory is improved.

[0021] Further, after adjusting the cleaning threshold, the prediction deviation value is determined based on the reacquired monitoring data. When the re-determined prediction deviation value is less than or equal to the preset prediction deviation value, the deviation between the predicted output and the actual output is within the qualified range after the increase of the complete monitoring frequency and the correction of the cleaning threshold. In this case, it is determined that the large output deviation is caused by the unqualified data prediction due to the abnormal sensor monitoring, and an abnormal alarm information of the high-risk sensor is issued. When the re-determined prediction deviation value is greater than the preset prediction deviation value, the prediction output deviation cannot be improved by improving the acquisition of monitoring data. In this case, it is determined that the actual output is abnormal due to the abnormal coke production, and the supervision parameter is corrected based on the output change representation value. Combined with the prediction deviation value and the high-risk sensor situation, the production management is comprehensively evaluated, and the supervision parameter is adjusted in time when the output deviation is large, so that the production process is continuously optimized, the production quality and efficiency are improved, and the monitoring efficiency for the intelligent factory is improved.

[0022] Further, the output change representation value reflects the change trend of the prediction deviation value. When the output change representation value is less than or equal to the preset output change representation value, the output deviation changes slowly, and in this case, the heat transfer is uneven and the reaction is insufficient in the coking process due to the abnormality of the coke production components, thereby reducing the total amount of coal charged into the furnace. At this time, the total amount of coal charged into the furnace is reduced to ensure that the unit mass of coal can obtain more sufficient heat under the same heating condition, so that the coal is heated more uniformly in the coking process, and the pyrolysis and carbonization reactions of the coal are ensured to proceed fully, thereby improving the coking effect. When the output change representation value is greater than the preset output change representation value, the prediction deviation value suddenly changes greatly, and in this case, it is determined that the raw material is abnormal, and an alarm information is issued. According to the output change trend, different measures are taken to improve the coking reaction conditions by reasonably adjusting the total amount of coal charged into the furnace, thereby improving the coke quality and output. The monitoring efficiency for the intelligent factory is improved.

[0023] Further, after completing the total amount of coal into the furnace adjustment, the prediction deviation value is determined based on the reacquired monitoring data, the ratio of the prediction deviation value before adjustment is calculated to obtain a change coefficient; when the change coefficient is less than or equal to a preset value, the yield prediction deviation is obviously improved, at this time, it is determined to continue to use the current coke generation parameters; when the change coefficient is greater than the preset change coefficient, the total amount of coal into the furnace adjustment does not completely solve the production problem, in this case, the coking time is adjusted to change the pyrolysis and carbonization time of coal, so that the reaction is more sufficient, and the coke quality and yield are improved. Further improve the monitoring efficiency of the intelligent factory. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 The step flow chart of the intelligent energy management method of the intelligent factory of the embodiment of the present application is shown in the figure. Figure 2 The module block diagram of the intelligent energy management system of the intelligent factory of the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0025] In order to make the purpose and advantages of the present application more clear and explicit, the present application will be further described below in combination with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present application, and are not used to limit the present application.

[0026] The preferred embodiments of the present application will be described below with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are only used to explain the technical principles of the present application, and are not used to limit the protection scope of the present application.

[0027] It should be noted that in the description of the present application, the terms indicating the direction or positional relationship of "up", "down", "left", "right", "inner", "outer" and the like are based on the direction or positional relationship shown in the drawings, which is only for the convenience of description, and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present application.

[0028] In addition, it should also be noted that in the description of the present application, unless otherwise explicitly specified and limited, the terms "mounting", "connecting", "connection" should be understood broadly, for example, it can be fixed connection, or detachable connection, or integrally connected; it can be mechanical connection, or electrical connection; it can be directly connected, or indirectly connected through an intermediate medium, or the communication inside two elements. Those skilled in the art can understand the specific meaning of the above terms in the present application according to the specific circumstances.

[0029] Please refer to Figure 1 The figure shows the step flow chart of the intelligent energy management method of the intelligent factory of the embodiment of the present application. The method described in the present application comprises: S1, monitoring each monitoring data in the coke preparation process, packaging the monitoring data of each sensor corresponding to the production single coke to obtain a plurality of monitoring data packets for each coke; S2, preprocessing the data of each monitoring data packet; S3, predicting the expected coke yield based on each monitoring data after cleaning; S4, determining whether the monitoring management for coke is qualified based on the prediction deviation value, comprising: determining that the monitoring management for coke is abnormal, adjusting the monitoring parameters for coke based on the detection data fluctuation parameter and the yield change characteristic value, including adjusting the monitoring frequency of each sensor to the corresponding value, adjusting the cleaning threshold used to filter out abnormal data in the preprocessing process to the corresponding value, adjusting the coking time in the coke refining process to the corresponding value or issuing corresponding alarm information; Or, determine that the monitoring management for coke is qualified, and continue to monitor the coke with the current monitoring parameters.

[0030] Specifically, the process of coke preparation is not limited, which can include proportioning various coking coals; coking various coking coals after proportioning into coke by high-temperature treatment; extracting the raw coal gas generated by the coking module and purifying the extracted raw coal gas; and cooling the coke output by the coking module.

[0031] Please refer to Figure 2 The module block diagram of the intelligent factory energy management system of the embodiment of the application is shown, and the system comprises: A coal blending module is used to proportion various coking coals; A coking module is connected with the coal blending module and is used to coking various coking coals after proportioning into coke by high-temperature treatment; A recovery module is connected with the coking module and is used to extract the raw coal gas generated by the coking module and purify the extracted raw coal gas; An extinguishing module is connected with the coking module and is used to cool the coke output by the coking module; A monitoring module is connected with the coal blending module, the coking module, the extinguishing module and the recovery module, and is used to obtain monitoring data of the coal blending module, the coking module, the extinguishing module and the recovery module; A data processing module is connected with the monitoring module and is used to preprocess each monitoring data; A prediction module is connected with the data processing module and is used to predict the expected coke yield based on each monitoring data after cleaning; an analysis module connected to the monitoring module and the prediction module respectively, configured to determine whether the monitoring management for the coke is qualified based on the predicted deviation value, and when it is determined that the monitoring management for the coke is abnormal, correct the monitoring parameters for the coke based on the detected data fluctuation parameter and the yield change characteristic value, including adjusting the monitoring frequency of the monitoring module to a corresponding value, adjusting the cleaning threshold of the data processing module to a corresponding value, adjusting the coking time of the coking module to a corresponding value, or issuing corresponding alarm information; an alarm module connected to the analysis module, configured to issue corresponding alarm information based on the determination result of the analysis module.

[0032] Specifically, the monitoring module includes an electronic belt scale sensor configured to obtain the total amount of coal charged into the furnace, a moisture meter configured to measure the moisture content of various coking coals, a vortex flowmeter configured to monitor the flow of heating coal gas, a thermocouple sensor configured to monitor the temperature of the combustion chamber, a thermal resistance sensor configured to monitor the temperature of the gas collecting pipe, a pressure transmitter configured to monitor the pressure of the gas collecting pipe, and a differential pressure transmitter configured to monitor the pressure at the bottom of the carbonization chamber. The speed sensor is arranged in the recovery module to monitor the speed of the air blower, the orifice plate flowmeter is arranged in the recovery module to monitor the flow of raw coal gas, and the belt scale is arranged at the output end of the coke quenching module to monitor the coke yield.

[0033] Specifically, the expected coke yield can be predicted by the prediction model.

[0034] Specifically, the training method of the prediction model and the algorithm of the model selection are not limited, and the gradient boosting tree (GBDT) can be used for training, which can be XGBoost, LightGBM or CatBoost. The training method of the model can align and resample all sensor data according to the timestamp, process missing values and abnormal values, determine the lag feature, and determine the timestamp of the data of each sensor corresponding to the predicted expected coke yield at a single time node. Obtain a plurality of sets of training data, and the data amount ratio of the training set to the validation set is 8:2. Each set of training data includes sensor data at each timestamp for predicting the expected coke yield and the actual coke yield. The training set data is used to fit the model to obtain a trained prediction model, and the input data of the prediction model is the preprocessed sensor monitoring data, and the output data is the expected coke yield.

[0035] Specifically, the monitoring data reflects the state of the production process, and by analyzing and modeling these data, the internal relationship between the production process and the coke yield is revealed. The prediction model learns the rules in the historical data, analyzes and predicts the current monitoring data, and obtains the expected coke yield.

[0036] Specifically, the data processing module further comprises a time locking unit, which is configured to package the monitoring data of each sensor corresponding to a single coke production furnace to obtain a monitoring data package.

[0037] Specifically, the process of determining whether the monitoring management of the coke is qualified based on the prediction deviation value comprises: The ratio of the absolute value of the difference between the expected coke production and the corresponding actually monitored coke production to the expected coke production is denoted as a prediction deviation value; When the prediction deviation value is less than or equal to a preset prediction deviation value, it is determined that the monitoring management of the coke is qualified, and the monitoring management of the coke is continued to be completed with the current monitoring parameters; When the prediction deviation value is greater than the preset prediction deviation value, it is determined that the monitoring management of the coke is abnormal, and the monitoring parameters for the coke are corrected based on the detection data fluctuation parameter.

[0038] Specifically, the preset prediction deviation value is selected within the interval [3%, 5%], and a person skilled in the art can determine the preset prediction deviation value according to actual application conditions. A large amount of historical production data can be statistically analyzed to calibrate the stable production process and the qualified product quality, and the preset prediction deviation value can be determined according to the analysis result. It can be understood that the case of whether the yield deviation is normal fluctuation can be divided. In the embodiment, the preset prediction deviation value is preferably 5%.

[0039] Specifically, the expected coke production of a single coke production furnace can be predicted based on the monitoring data package corresponding to the single coke production furnace, and the prediction deviation value can be determined according to the actual coke production and the expected coke production of the single coke production furnace to determine whether the monitoring management of the coke is qualified. It can be understood that a plurality of monitoring data packages can also be used to predict a plurality of coke production furnaces, and the prediction deviation value can be determined according to the total prediction result and the total actually monitored production.

[0040] Specifically, the process of determining the detection data fluctuation parameter comprises: A plurality of monitoring data of a single sensor for a single monitoring data package are obtained, the maximum value difference between the maximum value and the minimum value in each monitoring data is solved, the ratio of the maximum value difference to the average value of the obtained monitoring data is calculated to obtain a data fluctuation value of the single sensor for the single monitoring data package, and the ratio of the data fluctuation value to a preset data fluctuation value corresponding to the single sensor is calculated to obtain a fluctuation ratio value of the single sensor for the single monitoring data package. The average value of the fluctuation ratio values of the sensors of each data package used to determine the expected coke production is calculated to obtain the data fluctuation parameter.

[0041] Specifically, the specific manner of determining the preset data floating value corresponding to the single sensor is not limited, and the average of the data floating values of the single sensor of each data package corresponding to the coke determined to be qualified for monitoring management in the historical data can be determined as the preset data floating value of the corresponding sensor, which will not be repeated here.

[0042] Specifically, the data fluctuation parameter is determined, including determining the data floating value, the preset data floating value, and the floating ratio. The data floating value reflects the fluctuation degree of the single sensor data within a period of time. The preset data floating value represents the normal fluctuation range of the sensor. The floating ratio value measures the deviation degree of the single sensor data fluctuation from the normal range. The data fluctuation parameter comprehensively reflects the overall fluctuation of all the sensor data participating in the calculation.

[0043] Specifically, the process of correcting the monitoring parameter for coke based on the detected data fluctuation parameter includes: If the detected data fluctuation parameter is less than or equal to the preset detected data fluctuation parameter, the monitoring parameter for coke is corrected based on the yield change representation value; If the detected data fluctuation parameter is greater than the preset detected data fluctuation parameter, the monitoring frequency of the high-risk sensor is adjusted to the corresponding value based on the detected data fluctuation parameter.

[0044] Specifically, the preset detected data fluctuation parameter is selected within the interval [1.19, 1.26], and those skilled in the art can determine the preset detected data fluctuation parameter by themselves. It can be understood that the sensor monitoring data can be divided into abnormal fluctuation or not based on the analysis of the data fluctuation of each sensor in the production cycle qualified for monitoring management in the historical data. In the present embodiment, preferably, the preset detected data fluctuation parameter is 1.2.

[0045] Specifically, the prediction deviation value reflects the accuracy of the yield prediction and reflects the difference between the actual yield and the expected value. Based on the prediction deviation value, it is determined whether the monitoring management for coke is qualified. The prediction deviation value provides a quantitative basis for production management. When the prediction deviation value is greater than the preset prediction deviation value, the yield deviation is abnormal. In this case, the data fluctuation parameter is obtained, which comprehensively reflects the overall fluctuation of all the sensor data participating in the calculation. The fluctuation of the sensor monitoring data reflects the stability of the production process. When the detected data fluctuation parameter is greater than the preset detected data fluctuation parameter, the fluctuation is large. In this case, the data accuracy is abnormal due to sensor failure or environmental interference. For this case, the monitoring frequency of each sensor is adjusted to the corresponding value based on the detected data fluctuation parameter to increase the data analysis amount and improve the yield prediction accuracy, thereby improving the monitoring efficiency for the intelligent factory.

[0046] Specifically, the monitoring frequency of the high-risk sensor is adjusted to a corresponding value based on the detection data fluctuation parameter, wherein, The sensor with a floating comparison value greater than a preset floating comparison value is determined as a high-risk sensor. The increase range of the monitoring frequency of the high-risk sensor is positively correlated with the detection data fluctuation parameter.

[0047] Specifically, the preset floating comparison value is selected in the interval [1.34, 1.51], and a person skilled in the art can select and determine the preset floating comparison value. The distribution of the floating comparison values of each sensor in the historical data can be analyzed, and whether the sensor corresponding to the floating comparison value of each sensor appears abnormal can be calibrated. The preset floating comparison value is determined based on the distribution data of the sensor floating comparison value of the sensor appearing abnormal in the monitoring data. In the embodiment, preferably, the preset floating comparison value is 1.5.

[0048] Specifically, the sensor data fluctuation is large when the floating comparison value is high, and the sensor may be abnormal, and is determined as a high-risk sensor. The greater the detection data fluctuation parameter, the more serious the problem, and the monitoring frequency of the high-risk sensor is increased to ensure timely and accurate grasp of the production state. Concentrating resources to monitor the high-risk sensor can improve the probability of discovering problems. According to the fluctuation degree, the monitoring frequency is dynamically adjusted, the data for analysis is increased, the amount of accurate data is increased, the accuracy of data prediction is ensured, and the monitoring efficiency of the intelligent factory is improved.

[0049] In the embodiment, optionally, The detection data fluctuation parameter is compared with a first preset fluctuation comparison value and a second preset fluctuation comparison value; When the detection data fluctuation parameter is less than or equal to the first preset fluctuation comparison value, the monitoring frequency of each high-risk sensor is adjusted to 1.11 times of the corresponding initial monitoring frequency; When the detection data fluctuation parameter is less than or equal to the second preset fluctuation comparison value and greater than the first preset fluctuation comparison value, the monitoring frequency of each high-risk sensor is adjusted to 1.18 times of the corresponding initial monitoring frequency; When the detection data fluctuation parameter is greater than the second preset fluctuation comparison value, the monitoring frequency of each high-risk sensor is adjusted to 1.25 times of the corresponding initial monitoring frequency; The first preset fluctuation comparison value is 1.23B0, the second preset fluctuation comparison value is 1.31B0, and B0 is a preset detection data fluctuation parameter.

[0050] Specifically, when the adjustment of the monitoring frequency of each high-risk sensor is completed, the cleaning threshold of the data processing module for screening abnormal data is adjusted to a corresponding value based on the actual coke production of a single coke oven, wherein, The adjustment range of the cleaning threshold is positively correlated with the actual coke production of a single coke oven.

[0051] In this embodiment, optionally, The actual coke output of the single-furnace coke is compared with the first preset output and the second preset output. If the actual coke output of the single-furnace coke is less than or equal to the first preset output, the high critical value of the cleaning threshold is adjusted to 1.11 times the initial high critical value, and the low critical value of the cleaning threshold is adjusted to 0.89 times the initial low critical value. If the actual coke output of the single-furnace coke is less than or equal to the second preset output and greater than the first preset output, the high critical value of the cleaning threshold is adjusted to 1.18 times the initial high critical value, and the low critical value of the cleaning threshold is adjusted to 0.82 times the initial low critical value. If the actual coke output of the single-furnace coke is greater than the second preset output, the high critical value of the cleaning threshold is adjusted to 1.23 times the initial high critical value, and the low critical value of the cleaning threshold is adjusted to 0.77 times the initial low critical value. The first preset output is 1.32L0, and the second preset output is 1.61L0, where L0 is the average of the actual coke outputs of the furnaces in the historical data.

[0052] Specifically, after the high-risk sensor monitoring frequency adjustment is completed, the cleaning threshold of the data processing module for screening out abnormal data is adjusted to the corresponding value based on the actual output of the single-furnace coke, and the increase range of the cleaning threshold is positively correlated with the actual output of the single-furnace coke. The actual output of the single-furnace coke reflects the production situation. When the output is high, the data fluctuation is larger, and the cleaning threshold is increased to retain more effective data. When the output is high, the production process changes more, the data fluctuation is larger, and the screening threshold is increased to avoid misjudging effective data as abnormal. According to the output, the cleaning threshold is dynamically adjusted to improve the accuracy and effectiveness of data processing, so that the data processing strategy can adapt to different production states and better reflect the actual production situation. The monitoring efficiency of the intelligent factory is improved.

[0053] Specifically, the mean and the standard deviation of the historical data can be calculated, and the mean plus or minus the standard deviation of the preset cleaning multiple is used as the cleaning threshold range.

[0054] Specifically, the mean plus the standard deviation of the preset cleaning multiple can be determined as the high critical value, and the mean minus the standard deviation of the preset cleaning multiple can be determined as the low critical value.

[0055] The data processing module is used to traverse the monitoring data of each sensor, mark the data exceeding the cleaning threshold range as abnormal data, and remove it from the data set.

[0056] Specifically, when the adjustment of the cleaning threshold is completed, the prediction deviation value is determined based on the reacquired monitoring data, and whether the monitoring management for the coke is qualified is determined based on the re-determined prediction deviation value, including: When the re-determined prediction deviation value is less than or equal to the preset prediction deviation value, it is determined that the monitoring management of the coke is qualified, and an alarm information for each high-risk sensor abnormality is sent out; When the re-determined prediction deviation value is greater than the preset prediction deviation value, it is determined that the monitoring management of the coke is abnormal, and the monitoring parameter for the coke is corrected based on the yield change representation value.

[0057] Specifically, after the cleaning threshold value is adjusted, the prediction deviation value is determined based on the re-acquired monitoring data. When the re-determined prediction deviation value is less than or equal to the preset prediction deviation value, the deviation between the predicted yield and the actual yield is within the qualified range after the increase of the complete monitoring frequency and the correction of the cleaning threshold value, and in this case, it is determined that the yield deviation is caused by the unqualified data prediction due to the sensor monitoring abnormality. At this time, the high-risk sensor abnormality alarm information is sent out. When the re-determined prediction deviation value is greater than the preset prediction deviation value, the prediction yield deviation cannot be improved by improving the acquisition of the monitoring data, and in this case, it is determined that the actual yield is abnormal due to the abnormality of the coke production. In this case, the monitoring parameter is corrected based on the yield change representation value. In combination with the prediction deviation value and the high-risk sensor condition, the production management is comprehensively evaluated, the monitoring parameter is adjusted in time when the yield deviation is large, the production process is continuously optimized, the production quality and efficiency are improved, and thus the monitoring efficiency of the intelligent factory is improved.

[0058] Specifically, the process of correcting the monitoring parameter for the coke based on the yield change representation value includes: A prediction deviation value time domain curve is drawn based on the prediction deviation value in the acquired historical data, and a slope of the curve at a current time node is determined as a yield change representation value; When the yield change representation value is less than or equal to a preset yield change representation value, the total amount of the coal fed into the single coke furnace is adjusted to a corresponding value based on the prediction deviation value; When the yield change representation value is greater than the preset yield change representation value, an alarm information for the raw material abnormality is sent out; The reduction range of the total amount of the coal fed into the single coke furnace is positively correlated with the prediction deviation value.

[0059] In the embodiment, optionally, The prediction deviation value is compared with a first deviation comparison value and a second deviation comparison value; If the prediction deviation value is less than or equal to the first deviation comparison value, the total amount of the coal fed into the single coke furnace is adjusted to 0.93 times the initial total amount of the coal fed into the single coke furnace; If the prediction deviation value is less than or equal to the second deviation comparison value and greater than the first deviation comparison value, the total amount of the coal fed into the single coke furnace is adjusted to 0.84 times the initial total amount of the coal fed into the single coke furnace; If the predicted deviation value is greater than the second deviation comparison value, the total amount of coal into the single furnace is adjusted to 0.76 times the initial total amount of coal into the single furnace; The first deviation comparison value is 1.14C0, the second deviation comparison value is 1.23C0, and C0 is a preset predicted deviation value.

[0060] Specifically, the preset yield change representation value is selected within the interval [0.01, 0.02] with a unit of % / day. Those skilled in the art can determine the preset floating comparison value by themselves. The preset yield change representation value can be determined by statistical analysis of the slope of the time domain curve of the predicted deviation value in the historical data. It can be understood that the abnormal increase can be divided. In the embodiment, the preset yield change representation value is preferably 0.01.

[0061] Specifically, the yield change representation value reflects the change trend of the predicted deviation value. When the yield change representation value is less than or equal to the preset yield change representation value, the yield deviation changes slowly. In this case, due to the abnormality of the coke production components, the heat transfer is uneven and the reaction is insufficient in the coking process, thereby reducing the total amount of coal into the furnace. At this time, the total amount of coal into the furnace is reduced to ensure that under the same heating conditions, the unit mass of coal can obtain more sufficient heat, so that the coal is heated more uniformly in the coking process, and the pyrolysis and carbonization reactions of the coal are ensured to proceed fully, thereby improving the refining effect. When the yield change representation value is greater than the preset yield change representation value, the predicted deviation value suddenly changes greatly. In this case, it is determined that the abnormality is caused by the raw material. At this time, an alarm information is issued. Different measures are taken according to the yield change trend, the coking reaction conditions are improved by reasonably adjusting the total amount of coal into the furnace, and the coke quality and yield are improved. The monitoring efficiency for the intelligent factory is improved.

[0062] Specifically, when the adjustment of the total amount of coal into the furnace is completed, it is determined whether to modify the generation parameters for coke based on the change coefficient, including: The predicted deviation value is determined based on the reacquired monitoring data, the ratio of the predicted deviation value before the total amount of coal into the furnace is adjusted is solved, and the change coefficient is obtained; If the change coefficient is less than or equal to the preset change coefficient, it is determined that the current coke generation parameters are continuously used to complete the coking of the coke; If the change coefficient is greater than the preset change coefficient, the coking time in the coke coking process is adjusted to the corresponding value based on the predicted deviation value; The increase range of the coking time for coke is positively correlated with the predicted deviation value.

[0063] Specifically, the preset change coefficient is selected in the interval [0.93, 0.96], and those skilled in the art can select and determine the preset change coefficient according to the actual situation. It can be understood that the adjustment before and after the prediction deviation value can be divided into the case of obvious improvement. In the embodiment, preferably, the preset change coefficient is 0.93.

[0064] In the embodiment, optionally, The prediction deviation value is compared with the first deviation comparison value and the second deviation comparison value; If the prediction deviation value is less than or equal to the first deviation comparison value, the coking time for the coke is adjusted to 1.12 times of the initial coking time; If the prediction deviation value is less than or equal to the second deviation comparison value and greater than the first deviation comparison value, the coking time for the coke is adjusted to 1.14 times of the initial coking time; If the prediction deviation value is greater than the second deviation comparison value, the coking time for the coke is adjusted to 1.18 times of the initial coking time.

[0065] Specifically, after the total amount of the coal into the furnace is adjusted, the prediction deviation value is determined based on the reacquired monitoring data, and the change coefficient is obtained by calculating the ratio of the prediction deviation value before the adjustment. When the change coefficient is less than or equal to the preset value, the prediction deviation of the yield is obviously improved, and at this time, it is determined to continue to use the current coke generation parameter. When the change coefficient is greater than the preset change coefficient, the total amount of the coal into the furnace does not completely solve the production problem, and in this case, the coking time is adjusted to change the pyrolysis and carbonization time of the coal, so that the reaction is more sufficient, and the coke quality and yield are improved. Further, the monitoring efficiency for the intelligent factory is improved.

[0066] Thus, the technical solutions of the present application have been described in connection with the preferred embodiments illustrated in the drawings, but those skilled in the art can easily understand that the protection scope of the present application is obviously not limited to these specific embodiments. Those skilled in the art can make equivalent changes or replacements to the related technical features without departing from the principles of the present application, and the technical solutions after the changes or replacements will fall within the protection scope of the present application.

[0067] The above description is only the preferred embodiments of the present application and is not used to limit the present application; for those skilled in the art, the present application can have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. An intelligent energy management method for smart factories, characterized in that, include: Monitor various data during the coke preparation process, package the monitoring data of each sensor corresponding to the production of a single furnace of coke, and obtain several monitoring data packages for each furnace of coke. Preprocess the data from each monitoring data packet; Based on the monitoring data after cleaning is completed, the expected coke production is predicted. Determining the compliance of monitoring and management for coke based on prediction deviation values ​​includes: Identify monitoring and management anomalies for coke, and adjust the regulatory parameters for coke based on fluctuation parameters of detection data and characteristic values ​​of production changes. This includes adjusting the monitoring frequency of each sensor to the corresponding value, adjusting the cleaning threshold used to screen out abnormal data during preprocessing to the corresponding value, and adjusting the coking time in the coke refining process to the corresponding value or issuing corresponding alarm information. Alternatively, determine that the monitoring and management of coke is up to standard, and continue to monitor and manage coke using the current monitoring parameters.

2. The intelligent smart factory energy management method according to claim 1, characterized in that, The process of determining whether the monitoring and management of coke is up to standard based on the predicted deviation value includes: The absolute value of the difference between the expected coke production and the corresponding actual monitored coke production is recorded as the prediction deviation value. When the predicted deviation value is less than or equal to the preset predicted deviation value, the monitoring and management of coke is deemed qualified, and the monitoring and management of coke is continued to be completed with the current monitoring parameters. When the predicted deviation value is greater than the preset predicted deviation value, an anomaly is identified in the monitoring and management of coke, and the regulatory parameters for coke are corrected based on the fluctuation parameters of the detection data.

3. The intelligent smart factory energy management method according to claim 2, characterized in that, The process of determining the fluctuation parameters of the detection data includes: Acquire several monitoring data points for a single sensor for a single monitoring data packet, solve for the maximum and minimum value difference between the maximum and minimum values ​​in each monitoring data point, calculate the ratio of the maximum and minimum value difference to the average value of the acquired monitoring data points, obtain the data fluctuation value of a single sensor for a single monitoring data packet, calculate the ratio of the data fluctuation value to the preset data fluctuation value corresponding to the single sensor, and obtain the fluctuation comparison value of a single sensor for a single monitoring data packet. The average of the fluctuation comparison values ​​of each sensor for each data packet used to determine the expected coke production is calculated to obtain the data fluctuation parameter.

4. The intelligent smart factory energy management method according to claim 3, characterized in that, The process of adjusting regulatory parameters for coke based on fluctuations in detection data includes: If the fluctuation parameter of the detection data is less than or equal to the preset fluctuation parameter of the detection data, the regulatory parameters for coke will be adjusted based on the characteristic value of production change. If the fluctuation parameter of the detection data is greater than the preset fluctuation parameter of the detection data, the monitoring frequency of each sensor will be adjusted to the corresponding value based on the fluctuation parameter of the detection data.

5. The intelligent smart factory energy management method according to claim 4, characterized in that, Based on the fluctuation parameters of the detection data, the monitoring frequency of high-risk sensors is adjusted to the corresponding value, whereby... Sensors with floating comparison values ​​greater than preset floating comparison values ​​are identified as high-risk sensors. The increase in the monitoring frequency of high-risk sensors is positively correlated with the fluctuation parameter of the detection data.

6. The intelligent smart factory energy management method according to claim 5, characterized in that, When adjusting the monitoring frequency for each high-risk sensor, the cleaning threshold used by the data processing module to filter out abnormal data is adjusted to the corresponding value based on the actual coke production of a single furnace. The adjustment range of the cleaning threshold is positively correlated with the actual coke production of a single furnace.

7. The intelligent smart factory energy management method according to claim 6, characterized in that, When adjusting the cleaning threshold, the predicted deviation value is determined based on the reacquired monitoring data, and the compliance of the monitoring and management of coke is determined based on the re-determined predicted deviation value, including: When the redefined prediction deviation value is less than or equal to the preset prediction deviation value, the monitoring and management of coke is deemed qualified, and alarm information is issued for each high-risk sensor anomaly. When the redefined prediction deviation value is greater than the preset prediction deviation value, an anomaly is identified in the monitoring and management of coke, and the regulatory parameters for coke are corrected based on the production change characterization value.

8. The intelligent smart factory energy management method according to claim 7, characterized in that, The process of revising regulatory parameters for coke based on production change indicators includes: Based on the prediction deviation values ​​obtained from historical data, a time-domain curve of the prediction deviation value is plotted, and the slope of the curve at the current time node is determined as the characteristic value of output change. When the output change indicator value is less than or equal to the preset output change indicator value, the total amount of coal fed into the furnace for single-furnace coking is adjusted to the corresponding value based on the prediction deviation value. When the output change indicator value exceeds the preset output change indicator value, an alarm message will be issued for raw material anomalies. The decrease in the total amount of coal fed into the furnace is positively correlated with the prediction deviation.

9. The intelligent smart factory energy management method according to claim 8, characterized in that, When adjusting the total amount of coal fed into the furnace, the decision on whether to modify the coke production parameters is based on the change coefficient, including: Based on the newly acquired monitoring data, the predicted deviation value is determined, and its ratio to the predicted deviation value before adjusting the total amount of coal fed into the furnace is calculated to obtain the change coefficient. If the change coefficient is less than or equal to the preset change coefficient, it is determined that the current coke production parameters will continue to be used to complete the coke refining process. If the change coefficient is greater than the preset change coefficient, the coking time in the coke refining process will be adjusted to the corresponding value based on the predicted deviation value. The increase in coking time for coke is positively correlated with the prediction deviation.

10. An intelligent factory energy management system using the intelligent factory energy management method according to any one of claims 1-9, characterized in that, include: The coal blending module is used to blend various types of coke. The coking module, which is connected to the coal blending module, is used to refine the various cokes that have been blended into coke through high-temperature treatment. A recovery module, which is connected to the coking module, is used to extract the raw coal gas produced by the coking module and purify the extracted raw coal gas. A coke quenching module, which is connected to the coking module, is used to cool the coke output from the coking module. A monitoring module, which is connected to the coal blending module, the coking module, the coke quenching module and the recovery module, is used to acquire monitoring data for the coal blending module, the coking module, the coke quenching module and the recovery module; A data processing module, which is connected to the monitoring module, is used to preprocess the monitoring data. A prediction module, which is connected to the data processing module, is used to predict the expected coke production based on the monitoring data after the cleaning is completed. The analysis module, which is connected to the monitoring module and the prediction module respectively, is used to determine whether the monitoring and management of coke is qualified based on the prediction deviation value, and when it is determined that the monitoring and management of coke is abnormal, it corrects the regulatory parameters of coke based on the fluctuation parameters of the detection data and the characteristic value of the change in production, including adjusting the monitoring frequency of the monitoring module to the corresponding value, adjusting the cleaning threshold used by the data processing module to screen out abnormal data to the corresponding value, adjusting the coking time of the coking module to the corresponding value or issuing the corresponding alarm information; An alarm module, which is connected to the analysis module, is used to issue corresponding alarm information based on the judgment result of the analysis module.

Images