Search

CN-122022604-A - Coal blending planning method based on geometric constraint of reflectivity distribution curve of coal and rock

CN122022604ACN 122022604 ACN122022604 ACN 122022604ACN-122022604-A

Abstract

The invention provides a coal blending planning method based on geometric constraint of a coal-rock reflectivity distribution curve, and relates to the technical field of coking and coal blending; the method comprises the steps of extracting a coal rock reflectivity distribution curve of a historical successful coal blending scheme to form a historical successful curve cluster, further constructing a safe coal rock distribution envelope band, converting the safe coal rock distribution envelope band into additional constraint conditions in a planning model, constructing the planning model by taking the lowest cost of the blended coal as an objective function, and solving to obtain the optimal blended coal ratio. According to the invention, accurate constraint conditions are constructed through the full-waveform coal rock reflectivity distribution curve, so that the stability and prediction accuracy of coke quality are obviously improved, the minimization of the ton coke blending cost is realized, and the research and development period of new coal introduction or scheme adjustment is shortened.

Inventors

  • LUO YUEQING
  • LI XINRONG
  • AN ZIWEN

Assignees

  • 华院计算技术(上海)股份有限公司

Dates

Publication Date
20260512
Application Date
20260414

Claims (10)

  1. 1. The coal blending planning method based on geometric constraint of the reflectivity distribution curve of the coal rock is characterized by comprising the following steps of: s1, inputting a quality index of target coke; s2, screening all historical coal blending schemes with actual production results meeting quality indexes, and extracting corresponding reflectance distribution curves of the matched coal and rock to form historical successful curve clusters; s3, constructing a safe coal-rock distribution envelope band based on the historical success curve cluster, and converting the safe coal-rock distribution envelope band into an additional constraint condition in the planning problem; S4, constructing a planning model by taking the lowest cost of the matched coal as an objective function, wherein constraint conditions comprise constraint on the quality index of the matched coal and constraint on the proportioning index, and further comprise additional constraint conditions obtained in the step S3; and S5, solving the planning model to obtain the optimal matching coal ratio.
  2. 2. The method according to claim 1, wherein in the constraint condition of step S3, the constraint function is a coal-rock reflectivity profile of the blended coal.
  3. 3. The method according to claim 2, wherein the constraint construction method of step S3 is envelope constraints and/or similarity distance constraints, wherein, Constructing constraints by calculating statistical upper and lower boundaries of historical success curve clusters through envelope constraints; and calculating a central curve of the historical successful curve cluster by similarity distance constraint, and constructing constraint by limiting the distance between the reflectivity distribution curve of the matched coal and rock and the central curve.
  4. 4. A method according to claim 3, wherein the specific construction method of the envelope constraint is as follows: ; Wherein, the Is the mass percentage of the j-th single coal in the mixed coal, Is the reflectivity distribution curve vector of the j-th single coal and rock, , Representation of The portion within the kth reflectance interval, K being the total number of intervals; 、 The statistical lower boundary and upper boundary of the historical success curve cluster are in the kth reflectivity interval of the coal rock reflectivity distribution, and the statistical lower boundary and upper boundary of the historical success curve cluster comprise boundary scaling through a relaxation factor.
  5. 5. The method of claim 3, wherein the step of, The center curve is a mean curve; The distance is the square of the euclidean distance.
  6. 6. The method according to claim 5, wherein the specific construction method of the similarity distance constraint is as follows: ; Wherein, the Is the mass percentage of the j-th single coal in the mixed coal, Is the reflectivity distribution curve vector of the j-th single coal and rock, , Representation of The portion within the kth reflectance interval, K being the total number of intervals; mean curve of historical success curve cluster Is used to determine the (k) th component of the (c), Is the maximum allowable variance threshold.
  7. 7. The method according to claim 1, wherein in step S4, The cost of the blended coal is calculated by the price and the proportion of each single coal; the quality index of the blended coal is the content proportion of each chemical component in the blended coal, and is obtained by weighting calculation of each single coal in the blended coal; The constraint on the proportioning index is the constraint on the maximum mass percentage allowed by single coal under the stock and process conditions.
  8. 8. A method according to claim 3, wherein the solving method of step S5 comprises: Solving the situation of only adopting envelope curve constraint by adopting a simplex method; and solving by adopting a sequence quadratic programming or interior point method solver under the condition of adopting similarity distance constraint or further superposing a nonlinear coke quality prediction model as additional constraint in the step S3.
  9. 9. A computer device comprising a memory, a processor and a computer program stored on the memory, characterized in that the processor executes the computer program to carry out the steps of the method according to any one of claims 1-8.
  10. 10. A computer program product, characterized in that the computer program, when being executed by a processor, implements the steps of the method according to any one of claims 1-8.

Description

Coal blending planning method based on geometric constraint of reflectivity distribution curve of coal and rock Technical Field The invention relates to the technical field of coking and coal blending, in particular to a coal blending planning method based on geometric constraint of a coal-rock reflectivity distribution curve. Background The existing automatic coal blending system is mainly based on linear programming, and aims at meeting chemical components (ash Ad, sulfur Std, volatile Vdaf) and specific lithology indexesOn the premise of lowest cost. The prior art has the following defects: (1) Information loss of tradition The average value is only one, and the "coal mixing structure" of the blended coal (such as whether double peaks, notch depths and the like appear or not) cannot be reflected. Also blended coals of average reflectivity, the distribution of active and inert components may be quite different, resulting in coke quality (CSR/CRI) fluctuations. The method is particularly characterized in that the quality dimension is that the problem of average value trap and coke quality fluctuation caused by the existing single index model only depends on the average reflectivity of a vitrinite in the prior artThe statistical indexes are subjected to linear programming, and the technical defects of coal blending structures (such as inferior mixed coal with normal distribution and double-peak distribution or middle gap) with the same average value but distinct microscopic distribution morphology cannot be identified and distinguished. This defect causes distortion of the predictive model, which results in uncontrolled fluctuations in the cold strength (M40/M10) and in the thermal properties (CSR/CRI) of the coke produced. (2) Nonlinear relation is often presented between coke quality and single coal proportion, and the prediction precision of a simple linear superposition model is low. The method is particularly characterized in that the cost dimension is that the coal blending cost is high due to insufficient model prediction precision, and the method aims at that the existing linear weighting model can not accurately reflect complex nonlinear interaction in the coking process, so that enterprises are forced to reserve an excessive quality safety allowance in a coal blending scheme for ensuring qualified coke quality, namely high-price high-quality main coking coal or fat coal is excessively added, and the cost of raw material coal is high. (3) Experience is difficult to multiplex, and successful experience of the master is often hidden in a coal-rock view, and is difficult to quantify into a mathematical model. The method is characterized in that the efficiency and the energy-saving dimension are long in new scheme research and development period and high in small coke oven experiment energy consumption, the traditional coal blending method mainly depends on manual experience trial blending, and physical verification is required to be carried out through a small coke oven experiment (40 kg or larger scale) with high frequency, so that the coal blending scheme is long in adjustment period (usually 3-5 days), high in experiment energy consumption and huge in manpower and time cost. Disclosure of Invention Aiming at the defects existing in the prior art, the invention provides a coal blending planning method based on geometric constraint of a coal rock reflectivity distribution curve, in particular to an optimization method for converting microstructure characteristics (coal rock curve) of historical successful coal blending into mathematical constraint, which aims to solve the technical problems that: (1) By introducing the full-waveform coal rock reflectivity distribution curve as geometric constraint, the problem of information loss is solved, and the substantial similarity of a new coal blending scheme and a historical high-quality scheme is ensured from the microscopic lithofacies structure level, so that the stability and the prediction accuracy of coke quality are remarkably improved. (2) The accurate nonlinear safety envelop belt is constructed by utilizing successful experience of historical big data, and on the premise of ensuring that the coke quality is not lower than a target value (such as CSR > 65), the blending burning potential of low-price coal types (such as weakly caking coal and gas coal) is accurately excavated, and unnecessary mass surplus is eliminated, so that the minimization of the ton coke blending coal cost is realized on a quantitative level. (3) Through the fingerprint characteristics of coal and rock successfully blended by digital multiplexing histories, the scheme that more than 90% of theories meet conventional indexes but the microstructure is unreasonable is eliminated in the calculation stage, the required physical experiment times are greatly reduced, the research and development period of new coal introduction or scheme adjustment is shortened, and the energy consump