CN-121998659-A - Deviation tracing and quantitative evaluation method and device for coal sample preparation process

CN121998659ACN 121998659 ACN121998659 ACN 121998659ACN-121998659-A

Abstract

The invention relates to the technical field of coal quality detection and process control, and discloses a method and a device for tracing deviation and quantitatively evaluating the coal sample preparation process, wherein the method comprises the steps of standardizing a sample preparation process, constructing a deviation transfer directed acyclic graph and a Bayesian model, the accurate definition of error source is realized, can clearly distinguish error source from the sampling, sampling or chemical examination link, and the specific problem process in the accurate location sampling process. Meanwhile, the contribution degree of the deviation is quantitatively calculated by means of real-time data and probabilistic reasoning, a subjective qualitative judgment mode that a traditional dependent engineer observes the appearance of a sample is thoroughly abandoned, and a quantitative conclusion that the deviation affects an assay result is accurately output.

Inventors

GAO WEI
CHEN XIAHUA
ZHANG LEI
LI CHAO
CHEN FANFENG
GUO XIAOHU
ZHANG WENXIN
BAO JUN
TIAN HAO

Assignees

杭州华电双冠能源科技有限公司

Dates

Publication Date: 20260508
Application Date: 20251224

Claims (10)

1. The method for tracing and quantitatively evaluating deviation in the coal sample preparation process is characterized by comprising the following steps of: inputting preset coal sample preparation process data of a preset coal field operation rule into a preset sample preparation process model, and outputting a plurality of sample preparation process nodes and deviation influencing factors of the sample preparation process nodes; Setting deviation influence factors of all the sample preparation process nodes as father nodes and all the coal sample preparation process nodes as child nodes to construct a directed acyclic graph; Distributing a conditional probability table to each node of the directed acyclic graph, and constructing a sample preparation deviation Bayesian model based on all the nodes and the corresponding conditional probability tables; inputting the real-time coal sample preparation process data into the sample preparation deviation Bayesian model, and evaluating the real-time coal sample preparation process through the sample preparation deviation Bayesian model to obtain an evaluation result.
2. The method of claim 1, wherein inputting the coal sample process data of the preset coal field operation protocol into the preset sample process model, outputting a plurality of sample process nodes and deviation influencing factors of each of the sample process nodes, comprises: acquiring a coal field operation rule; Inputting the coal sample preparation process data of the coal field operation rules into a preset sample preparation flow model; Analyzing the whole sample preparation process of the coal sample preparation process data through the sample preparation process model to obtain a series of continuous sample preparation process nodes; And analyzing deviation influence factors of each sample preparation process node deviating from ideal conditions.
3. The method of claim 1, wherein constructing the directed acyclic graph with the bias influencing factors for each of the sample preparation process nodes as parent nodes and each of the coal sample preparation process nodes as child nodes comprises: Setting all deviation influencing factors of each sample preparation process node as father nodes, and setting each sample preparation process node as child nodes; And connecting the father node and the child nodes corresponding to the father node by using directional arrows to obtain a directional acyclic graph.
4. The method of claim 1, wherein the assigning conditional probability tables to each node of the directed acyclic graph and constructing a sample bias bayesian model based on all the nodes and corresponding conditional probability tables comprises: defining node states of all nodes of the directed acyclic graph; Constructing a conditional probability table of each node based on node state combinations of the parent nodes and node state probability distributions of the corresponding child nodes; determining the conditional probability parameters in the conditional probability table of each node according to the historical operation data, the equipment design tolerance, the experimental test data and the expert experience; and constructing a sample preparation deviation Bayesian model by adopting each node and the conditional probability table corresponding to each node.
5. The method of claim 1, wherein inputting the real-time coal sample preparation process data into the sample preparation bias bayesian model, evaluating the real-time coal sample preparation process through the sample preparation bias bayesian model to obtain an evaluation result, and comprising: acquiring a real-time current value, equipment state parameters, operation parameters and intermediate quality parameters in a real-time coal sample preparation process; performing data processing on the real-time current value, the equipment state parameter, the operation parameter and the intermediate quality parameter and inputting the sample preparation deviation Bayesian model; Loading the nodes and the corresponding conditional probability tables through the sample preparation deviation Bayesian model, and converting the real-time current value, the equipment state parameter, the operation parameter and the intermediate quality parameter after data processing into discrete state evidence which can be identified by the sample preparation deviation Bayesian model; Compiling the sample preparation deviation Bayesian model by adopting a preset Bayesian algorithm, and inputting all the discrete state evidences into the compiled sample preparation deviation Bayesian model to carry out global probability propagation; Updating probability distribution of unknown nodes of the compiled sampling deviation Bayesian model according to the propagation result; Extracting posterior probability distribution of all nodes of the updated sample preparation deviation Bayesian model, and screening posterior probability distribution of preset target ash deviation nodes; Extracting posterior probability values corresponding to states of all nodes from posterior probability distribution of the target ash deviation nodes, and selecting a maximum posterior probability value; Determining the optimal state and deviation evaluation information of the target ash deviation node based on the maximum posterior probability value; comparing the prior probability value and the posterior probability value of each node, and determining the node corresponding to the key deviation influence factor according to the comparison result; calculating the difference value between the prior probability value and the posterior probability value of each node to obtain the contribution degree of each node; and generating an evaluation result by adopting deviation evaluation information of the target ash deviation node, the node corresponding to the key deviation influence factor and the contribution degree of each node.
6. The method according to claim 1, wherein the method further comprises: sending the evaluation result to an operation interface or a centralized control system, and generating an inspection work order matched with the evaluation result; and executing predictive maintenance operation based on the checking work order.
7. The utility model provides a coal sample preparation process deviation traceability and quantitative evaluation device which characterized in that, the device includes: the input module is used for inputting preset coal sample preparation process data of a preset coal field operation rule into a preset sample preparation flow model and outputting a plurality of sample preparation process nodes and deviation influence factors of the sample preparation process nodes; the construction module is used for constructing a directed acyclic graph by taking deviation influence factors of all the sample preparation process nodes as father nodes and all the coal sample preparation process nodes as child nodes; the distribution module is used for distributing a conditional probability table to each node of the directed acyclic graph, and constructing a sample preparation deviation Bayesian model based on all the nodes and the corresponding conditional probability tables; the evaluation module is used for inputting the real-time coal sample preparation process data into the sample preparation deviation Bayesian model, and evaluating the real-time coal sample preparation process through the sample preparation deviation Bayesian model to obtain an evaluation result.
8. An electronic device, comprising: The coal sample preparation process deviation tracing and quantitative evaluation method comprises a memory and a processor, wherein the memory and the processor are in communication connection, the memory stores computer instructions, and the processor executes the computer instructions, so that the coal sample preparation process deviation tracing and quantitative evaluation method is executed.
9. A computer-readable storage medium having stored thereon computer instructions for causing a computer to perform the method of provenance and quantitative assessment of coal sample production process bias of any of claims 1 to 6.
10. A computer program product comprising computer instructions for causing a computer to perform the method of provoking and quantifying deviations in coal-sampling processes of any of claims 1-6.

Description

Deviation tracing and quantitative evaluation method and device for coal sample preparation process Technical Field The invention relates to the technical field of coal quality detection and process control, in particular to a deviation tracing and quantitative evaluation method and device for a coal sample preparation process. Background The coal sample preparation is a key core link of connection sampling and assay in a coal quality detection system, and an original coal sample is prepared into a representative analysis sample through crushing, mixing, shrinkage dividing, drying and other series of operations, so that a foundation is provided for the detection of key quality indexes such as subsequent heat value, sulfur content, ash content and the like. The accuracy of the coal quality index is directly related to trade settlement fairness and production coal blending scientificity, and sample preparation deviation can be transmitted and amplified to a final test result, so that trade disputes are caused, or the problems of coal blending decision errors, energy consumption increase and the like are caused. Therefore, the standardization and reliability of coal sample preparation are important to the efficient operation of the whole industrial chain of coal, and are the precondition for guaranteeing the accuracy of coal quality detection. Therefore, the detection and evaluation of the sample preparation quality generally mainly adopts a result guiding and controlling method, specifically only basic result data such as the weight of raw coal samples, the sample preparation time and the like are recorded, and whether the sample preparation quality is qualified is reversely deduced through a final test result. However, the method lacks traceability and quantitative evaluation capability, the specific operation states of the crushing, shrinkage and other procedures of each stage cannot be completely restored in the existing records, when the test result is abnormal, it is difficult to accurately judge whether an error source is a sample preparation, sampling or test link, a specific problem procedure cannot be positioned even if the sample preparation link is locked, qualitative judgment is carried out by observing appearance characteristics such as granularity, humidity and the like of a sample by an engineer, subjectivity is strong, and a quantitative conclusion of influence of deviation on the test result cannot be given. Disclosure of Invention The invention provides a deviation tracing and quantitative evaluation method and device in a coal sample preparation process, which are used for solving the problems that whether an error source is a sample preparation, sampling or testing link is difficult to accurately judge, a specific problem process cannot be positioned even if the sample preparation link is locked, and an engineer is relied on to observe appearance characteristics such as sample granularity, humidity and the like to carry out qualitative judgment, so that subjectivity is strong, and a quantitative conclusion of influence of deviation on a testing result cannot be given. In a first aspect, the invention provides a method for tracing and quantitatively evaluating deviation of a coal sample preparation process, which comprises the following steps: inputting preset coal sample preparation process data of a preset coal field operation rule into a preset sample preparation process model, and outputting a plurality of sample preparation process nodes and deviation influencing factors of the sample preparation process nodes; Setting deviation influence factors of all the sample preparation process nodes as father nodes and all the coal sample preparation process nodes as child nodes to construct a directed acyclic graph; Distributing a conditional probability table to each node of the directed acyclic graph, and constructing a sample preparation deviation Bayesian model based on all the nodes and the corresponding conditional probability tables; inputting the real-time coal sample preparation process data into the sample preparation deviation Bayesian model, and evaluating the real-time coal sample preparation process through the sample preparation deviation Bayesian model to obtain an evaluation result. According to the invention, through standardized disassembly of the sample preparation flow, construction of the deviation transfer directed acyclic graph and the Bayesian model, accurate definition of error sources is realized, the error sources can be clearly distinguished from sample preparation, sampling or testing links, and specific problem procedures in the sample preparation process can be accurately positioned. Meanwhile, the contribution degree of the deviation is quantitatively calculated by means of real-time data and probabilistic reasoning, a subjective qualitative judgment mode that a traditional dependent engineer observes the appearance of a sample is thoroughly abandoned, and a qua