Container coke quality early warning method based on bulk density analysis

Abstract

The invention discloses a bulk density analysis-based container coke quality early warning method, which comprises the following steps of: acquiring data such as container size, weighing gross weight / tare weight, historical net weight of suppliers, size fraction and the like, calculating volume and bulk density of a container body, counting average net weight in a short period of time according to the suppliers, and generating a net weight early warning value; the method comprises the following steps: calculating an actually measured net weight in real time and comparing the actually measured net weight with an early warning value during factory weighing, automatically giving an alarm when the net weight exceeds a threshold value, classifying and grouping, issuing targeted sampling detection, writing a detection result back to a database to correct an early warning model, and forming an alarm-detection-feedback closed loop. Full-coverage rapid primary screening can be achieved for each box in the weighing link, the problems of insufficient coverage and detection lag of traditional sampling are remarkably solved, granularity abnormity can be recognized more accurately, suppliers are prevented from avoiding inspection through a boxing means, the response time is shortened, and the quality risk and the inspection cost are reduced. The method is suitable for a container coke scene with full container delivery as a main part, and early warning parameters can be dynamically adjusted according to historical data or a more complex model.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of steel smelting raw material detection, and particularly relates to a containerized coke quality early warning method based on bulk density analysis. BACKGROUND

[0002] Coke particle size is crucial to blast furnace production, and is directly related to the permeability, reaction efficiency, heat distribution and stable operation of the blast furnace. For steel enterprises without coking production, coke particle size detection is an important part of coke quality detection, and the accuracy of the particle size detection directly affects the stability of the blast furnace production. For containerized coke, due to the manpower and equipment level, it is impossible to detect every container upon entering the factory, and only sampling detection can be used for detection.

[0003] The existing sampling detection method has several shortcomings and risks that cannot be ignored: the quality of the coke may fluctuate within the batch or be stratified during production, packaging, transportation and stacking, random sampling cannot cover all risky containers, and there may be a situation of "covering up the inferior with the superior". If a small number of containers contain inferior coke with small particle size or high powder, the sampling probability is low, but once used in the blast furnace, it may cause deterioration of the blast furnace permeability or fluctuation of the furnace condition, resulting in significant production and economic losses; traditional physical and chemical detection (such as thermal performance and industrial analysis) usually takes several hours or even several days, and the detection results cannot be fed back to the receiving site in real time, making it difficult to identify and dispose of the coke upon entering the factory; the fixed or regular sampling process may be used by suppliers or carriers (for example, by stratified loading during packaging), increasing the risk of "substituting inferior for superior" and weakening the deterrent effect of quality control. SUMMARY

[0004] In order to overcome the above-mentioned shortcomings of the prior art, the purpose of the present application is to provide a containerized coke quality early warning method based on bulk density analysis.

[0005] The technical solution adopted by the present application to solve its technical problems is: a containerized coke quality early warning method based on bulk density analysis, comprising the following steps:

[0006] S1: Collect and store the supplier identification, container number, container size, single box gross weight, last tare weight, historical net weight, particle size and detection item data of the containerized coke, and establish a big data analysis database;

[0007] S2: Based on the database, calculate the average net weight of the single box within 30 days according to the supplier, and generate the container net weight early warning value of each supplier according to the formula based on the correlation between the container volume, coke particle size and bulk density;

[0008] S3, parameterize the early warning value;

[0009] S4, write the early warning value of each supplier into the enterprise intelligent management and control / metering quality system, and connect with the real-time data of the weighbridge; when the container enters the factory and is weighed by the weighbridge to obtain the gross weight, the system reads the corresponding tare weight of the nearest time and calculates the measured net weight, and compares the measured net weight with the early warning value corresponding to the supplier;

[0010] S5, alarm triggering and information pushing: when the measured net weight-early warning value>0, or the preset alarm condition is met, the system automatically generates alarm information and pushes the alarm information through a report;

[0011] S6, the quality supervision and acceptance department classifies and batches the containers triggering the early warning, and implements full coverage or targeted sampling detection on the classified and batched containers;

[0012] S7, according to the targeted detection result, the early warning container is disposed as qualified or unqualified, and the detection and disposal record is fed back and updated to correct.

[0013] As a further improvement of the application: the early warning value is determined according to the following formula: early warning value=average net weight of single container x correlation coefficient K, wherein the correlation coefficient K is a preset value.

[0014] As a further improvement of the application: the correlation coefficient K is in the range of 1.05-1.2. Preferably, the correlation coefficient K is 1.1.

[0015] As a further improvement of the application: the container size collected in step S1 is used to calculate the volume of the container, and a statistical correlation model between the coke bulk density and the particle size is calculated and established.

[0016] As a further improvement of the application: the tare weight used for judgment in step S4 is the nearest recorded tare weight of the same car number / container; when there is no nearest tare weight, the system uses a preset tare weight or suspends automatic judgment and prompts manual verification according to the rules.

[0017] As a further improvement of the application: the alarm information in step S5 is connected to the remote metering system through the sailsoft report pushing system; the alarm information includes car number, manufacturer and gross weight time.

[0018] As a further improvement of the application: the batching rule of the early warning container in step S6 includes: batching according to contract number, manufacturer and factory time.

[0019] As a further improvement of the application: in step S6, 10 cars are randomly selected for sampling per 10 cars / containers, and ≤8 cars are not counted as the number of groups; when the planned quantity is ≤16 cars, sampling is carried out immediately after unloading; the batching period is one batch every 48 hours.

[0020] As a further improvement of the present application: the targeted detection items in step S6 are: particle size, moisture, industrial analysis, drum test, thermal performance test, and bulk specific gravity test; and the historical particle size distribution and net weight per box statistics of the supplier are updated according to the detection results.

[0021] Compared with the prior art, the present application has the following beneficial effects:

[0022] The coke transported by containers is pre-judged in terms of particle size index by the system at the weighing link, the coke quality detection method is increased, and the influence of the non-uniformity of the sample caused by coke sampling, the amplification of small probability defects, the loopholes in the randomness of sampling, and the lag of the detection results on the coke particle size quality is improved.

[0023] The preliminary detection of the coke is carried out by using the early warning value at the coke weighing link in the plant, the coke sampling defects are improved, and the preliminary prediction result of the coke particle size quality is timely;

[0024] The early warning threshold value is used to realize the full coverage prediction, the sampling loopholes are compensated, the real-time monitoring and early warning are realized, the detection lag problem is avoided, the abnormal coke is accurately identified based on the big data analysis, the cheating behavior of the supplier is prevented, and the quality risk is reduced. DETAILED DESCRIPTION

[0025] In order to make the purpose, technical scheme and advantages of the present application clearer, the technical scheme of the present application will be described clearly and completely in combination with the specific embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0026] It should be noted that the terms "first", "second" and the like in the specification and claims of the present application are used to distinguish similar objects, and do not necessarily indicate a specific order or chronological sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not necessarily limit to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0027] The quality control process mainly based on sampling inspection: according to the contract number / manufacturer / arrival time, batches are grouped (usually every 48 hours), and a certain number of boxes are randomly selected for laboratory testing within each batch; the on-site weighing (weighbridge) and the measurement system are used to record the gross weight / tare weight, but most enterprises do not conduct immediate full inspection at the arrival stage, but rely on subsequent sampling results to determine whether the entire batch is qualified; some enterprises have experience-based sorting / re-inspection processes, and implement centralized sampling and more in-depth testing on batches that trigger abnormalities. Sampling only covers a certain number of boxes within the batch, and cannot guarantee the detection of scattered or small amounts of poor quality boxes. A small number of poor quality boxes can enter production, causing local blockage of the blast furnace, reduced permeability, or fluctuations in furnace conditions. If the single-box powder content is abnormally high, but the probability of being sampled is low, the production risk is often caused when detected. If the single-box powder content is abnormally high, but the probability of being sampled is low, the production risk is often caused when detected. Many key tests (such as thermal performance and industrial analysis) require a long time and cannot be immediately determined at the time of arrival / storage. Fixed or predictable sampling rules can be avoided by the packer / carrier through stratified loading, selecting packing time, etc. The existing process relies more on direct physical detection and less on using available weighbridge, box size, etc. data for real-time quality inference. Measurement data, supplier historical records, and test results are often scattered in different systems or manual accounts, and lack of automatic linkage and online early warning.

[0028] In order to solve the defects of existing sampling inspection, by the correlation between coke particle size and bulk density, an early warning mechanism is established to realize rapid prediction and full coverage detection of containerized coke at the time of arrival, and to ensure that the coke quality meets the requirements of blast furnace production.

[0029] The present application utilizes the principle that the smaller the coke particle size, the greater the bulk density, and establishes the net weight early warning value of the incoming containerized coke by analyzing the coke and coke fines particle size quality and the net weight of the containerized coke of the coking manufacturer that adopts containerized transportation. The early warning threshold is embedded in the intelligent management and control system to monitor the weighing data of the incoming containerized coke in real time. When the containerized coke exceeds the early warning value during weighing by the weighbridge, the early warning coke container is uniformly classified and batched by the coke quality supervision and acceptance department, and full coverage centralized sampling and detection are carried out. The detection is targeted, the coke quality detection method is refined, the defects of sampling inspection are improved, and the accuracy of coke quality detection is ensured.

[0030] The present application will be further described in conjunction with the embodiments: a containerized coke quality early warning method based on bulk density analysis, characterized in that it comprises the following steps:

[0031] S1: Collect and store the supplier identification, container number, container size, single-box gross weight, last tare weight, historical net weight, particle size, and detection item data of the containerized coke, and establish a big data analysis database;

[0032] S2: Based on the database, the average net weight of a single container of a supplier in the past 30 days is counted, the correlation between the volume of the container, the particle size of the coke and the bulk density is combined, and the net weight early warning value of each supplier is generated according to the formula;

[0033] S3, parameterization of early warning value;

[0034] S4, write the early warning value of each supplier into the enterprise intelligent management and control / metering quality system, and connect with the real-time data of the weighbridge; when the container enters the factory and is weighed by the weighbridge to obtain the gross weight, the system reads the corresponding tare weight and calculates the measured net weight, and compares the measured net weight with the early warning value corresponding to the supplier;

[0035] S5, alarm triggering and information pushing: when the measured net weight-early warning value>0, or the preset alarm condition is met, the system automatically generates an alarm information and pushes it through a report;

[0036] S6, the quality supervision and acceptance department classifies and batches the containers triggering the early warning, and implements full coverage or targeted sampling detection on the classified and batched containers;

[0037] S7, according to the targeted detection results, the early warning containers are disposed as qualified or unqualified, and the detection and disposal records are fed back and updated to correct.

[0038] Specifically, the present application utilizes the principle that the smaller the coke particle size, the greater the bulk density, and aims at the coke and coke dust particle size quality and the net weight of the container of the coking plant adopting container transportation, establishes the net weight early warning value of the container coke entering the factory through big data analysis, and the specific steps are as follows:

[0039] I. Data collection and analysis:

[0040] 1. Establish a big data analysis database to analyze and count the net weight of single container coke of each coking plant adopting container transportation;

[0041] 2. According to the database, the quality of the coke particle size of each plant is statistically analyzed;

[0042] 3. Through big data analysis, the net weight of single coke container corresponding to the quality of coke particle size of each plant is analyzed, the principle that the smaller the coke particle size, the greater the bulk density is utilized, the correlation between the quality of coke particle size and the net weight after the coke container is weighed is analyzed, the correlation between the coke net weight and the coke net weight early warning value is set, and the coke net weight early warning value of the coke entering the factory by container transportation is established;

[0043] II. Early warning system development:

[0044] 1. It is included in the intelligent management and control metering quality system of the steel enterprise, the program is written, the early warning function is developed, and the container coke net weight early warning value is set;

[0045] 2. The coke of the coking plant transported by the container is weighed by the ground scale when entering the plant, the intelligent management and control quality system is identified and analyzed, the container coke box body exceeding the early warning value is automatically edited by the system, and the information is sent to the relevant staff;

[0046] III. Targeted detection:

[0047] The coke quality supervision and acceptance department classifies and groups the early warning coke containers, conducts full-coverage centralized sampling detection, targeted detection, refines the coke quality detection method, and ensures the accuracy of coke quality detection and timely identification and processing of poor quality coke.

[0048] As a specific embodiment of the present application, a big data analysis database is established to analyze and count the net weight of single container coke of each coking plant adopting container transportation.

[0049]

[0050] Table 1 Big data statistics of average net weight of container coke entering the plant According to the database, the quality of the coke particle size of each plant is statistically analyzed;

[0051]

[0052] Table 2 Container size of each coking plant

[0053]

[0054] Table 3 Quality of coke particle size of each plant and container coke bulk density According to the big data analysis of the average net weight and the maximum net weight of the single coke container corresponding to the quality of the coke particle size of each plant, using the principle that the smaller the coke particle size, the greater the bulk density, the quality of the coke particle size is set to be related to the net weight after the container coke is weighed, and the coke particle size of each coking plant is set to be related to the net weight after the container coke is weighed.

[0055] According to the big data analysis, the average net weight of the container coke entering the plant of each coking plant is related to the correlation coefficient 1.1, and the net weight early warning value of the coke transported by the container entering the plant is established. The container coke that does not meet the requirements is identified, and the coke particle size quality that does not trigger the early warning can meet the contract standard and meet the stable production demand of the blast furnace.

[0056]

[0057] Table 4 Establishment of early warning value of container coke entering the plant of each coking plant

[0058] It is included in the intelligent management and control quality system of the steel enterprise, a program is written, a warning function is developed, and the coke early warning value of the coke entering the plant is set.

[0059] The company's report pushing system is connected with the company's remote metering system by using the company's sail soft report pushing system, the current gross weight time, the gross weight, the corresponding supplier production plant and the last time history tare weight of the vehicle are obtained, after the system obtains the data, the average net weight of the vehicle in the last month is obtained according to the vehicle number and the manufacturer information, and the net weight is 1.1, as a net weight warning value, when the gross weight weight - the last time tare weight - the warning net weight = the alarm value, if the alarm value is greater than zero, the alarm is given, the report pushing system is triggered to realize data pushing, and the data is pushed to the company's enterprise WeChat, and the business personnel with the permission will receive the notification.

[0060] When the coking plant coke transported by containers enters the plant and is weighed by a platform scale, the intelligent management and control quality system identifies and analyzes, and the container coke box body exceeding the early warning value is automatically edited by the system and sent to the relevant staff;

[0061] The coke quality supervision and acceptance department classifies and groups the early warning coke containers uniformly, carries out centralized sampling detection, carries out targeted detection, refines the coke quality detection method, ensures the accuracy of coke quality detection, and also eliminates the false behavior of coke suppliers (manufacturers).

[0062] The main functions of the present application are:

[0063] The present application introduces a big data "early warning value" mechanism based on bulk density / net weight-volume correlation in the container coke weighing link, carries out full coverage and rapid preliminary screening for each box, and is linked and disposed with traditional sampling detection, so that the timeliness, coverage rate, accuracy and anti-cheating ability are significantly improved in coke particle size quality control.

[0064] In the weighing link, the "early warning value" is used to identify each coke box in real time, which makes up for the blind area of traditional sampling which only covers a small number of boxes, and significantly reduces the risk of "covering up the inferior with the superior".

[0065] The weighing is identified and real-time alarm is given, which solves the problem of delayed response caused by laboratory detection lag, so that quality control can take measures at the receiving link.

[0066] Based on the statistical correlation of supplier historical net weight, box volume and particle size / bulk density, a customized early warning threshold is constructed, which can more accurately identify the box with abnormal powder content or fine particle size, and reduce the false alarm and omission rate.

[0067] Using the objective data indexes of weight-volume-bulk density, it is difficult for the supplier to systematically avoid sampling by layering or loading strategy, which fundamentally suppresses the behavior of "using inferior goods to replace good goods".

[0068] The boxes triggering the early warning are classified and grouped, and targeted sampling detection is carried out, the detection results are written back to the database for correction of the early warning model, realizing the closed-loop management of alarm-detection-correction, and improving the long-term quality control effect.

[0069] By early detection and targeted disposal of high-risk boxes, the cost of abnormal furnace conditions, production stoppage or rework caused by undetected poor coke is reduced, thereby improving the utilization of inspection resources while ensuring quality.

[0070] The early warning threshold, statistical window and determination rule can be adjusted according to historical data, supplier performance and actual site or automatically optimized by a model, facilitating integration with existing measurement systems of an enterprise and engineering deployment.

[0071] In summary, after reading the present application document, those skilled in the art can make other various corresponding transformation schemes according to the technical solutions and technical concepts of the present application without creative mental labor, which all belong to the scope of protection of the present application.

Claims

1. A method for early warning of containerized coke quality based on bulk density analysis, characterized in that, Includes the following steps: S1: Collect and store supplier identification, container number, container size, gross weight per container, most recent tare weight, historical net weight, particle size and test item data for coke transported by container, and establish a big data analysis database. S2: Based on the database, the average net weight of a single container within 30 days is calculated according to the supplier. Combining the correlation between container volume and coke particle size and bulk density, a container net weight warning value for each supplier is generated according to the formula. S3, Parameterization of early warning values; S4. Write the warning values ​​of each supplier into the enterprise's intelligent management / measuring quality system and connect it with the real-time data of the weighbridge; when the container enters the factory and the gross weight is obtained by weighing on the weighbridge, the system reads the most recent tare weight and calculates the actual net weight, and compares the actual net weight with the warning value corresponding to the supplier. S5. Alarm Triggering and Information Push: When the measured net weight minus the warning value > 0, or when the preset alarm conditions are met, the system automatically generates alarm information and pushes it through the report. S6. The quality supervision and acceptance department shall classify and group the containers that trigger the warning in a unified manner, and conduct full-coverage or targeted sampling and testing on the classified and grouped containers. S7. Based on the targeted inspection results, handle the early warning containers as qualified or unqualified, and provide feedback on the inspection and handling records and update the database for correction.

2. The method for early warning of containerized coke quality based on bulk density analysis according to claim 1, characterized in that, The warning value is determined by the following formula: Warning value = Average net weight per box × Correlation coefficient K, where the correlation coefficient K is a preset value.

3. The method for early warning of containerized coke quality based on bulk density analysis according to claim 2, characterized in that, The correlation coefficient K ranges from 1.05 to 1.

2.

4. The method for early warning of containerized coke quality based on bulk density analysis according to claim 3, characterized in that, The correlation coefficient K is 1.

1.

5. The method for early warning of containerized coke quality based on bulk density analysis according to claim 1, characterized in that, The container dimensions collected in step S1 are used to calculate the container volume, and to calculate and establish a statistical correlation model between coke bulk density and particle size.

6. The method for early warning of containerized coke quality based on bulk density analysis according to claim 1, characterized in that, In step S4, the tare weight used for identification is the most recently recorded tare weight for the same vehicle number / container. If no most recently recorded tare weight is available, the system uses a preset tare weight or pauses automatic identification according to the rules and prompts for manual verification.

7. The method for early warning of containerized coke quality based on bulk density analysis according to claim 1, characterized in that, In step S5, the alarm information is transmitted to the remote metering system via the FineReport system; the alarm information includes vehicle number, manufacturer, gross weight, and time.

8. The method for early warning of containerized coke quality based on bulk density analysis according to claim 1, characterized in that, The batching rules for the early warning containers in step S6 include: batching by contract number, manufacturer, and arrival time.

9. A method for early warning of containerized coke quality based on bulk density analysis according to claim 8, characterized in that, In step S6, 10 trucks are randomly selected for sampling in groups of 10 trucks / containers. If there are ≤8 trucks, the number of groups is not counted. If the planned quantity is ≤16 trucks, sampling is carried out immediately after unloading. The batching cycle is 48 hours per batch.

10. A method for early warning of containerized coke quality based on bulk density analysis according to claim 9, characterized in that, The targeted testing items mentioned in step S6 are: particle size, moisture content, industrial analysis, drum inspection, thermal performance test, and bulk density test; and the supplier's historical particle size distribution and single-box net weight statistics are updated based on the test results.

Images