CN-121974158-A - Pneumatic ash conveying system and operation and maintenance control method

CN121974158ACN 121974158 ACN121974158 ACN 121974158ACN-121974158-A

Abstract

The invention discloses a pneumatic ash conveying operation and maintenance control method, which relates to the technical field of pneumatic ash conveying and comprises the following steps of S1, intelligent sensing and digital modeling, namely collecting operation data and material characteristic data of a pneumatic ash conveying core component through distributed sensing nodes, constructing a digital twin body virtual mapping and standardized operation and maintenance database by combining federal learning, S2, multi-agent cooperative regulation and control, namely generating an optimal self-adaptive regulation and control strategy through multi-agent real-time interaction based on the virtual mapping and the database of S1, synchronously executing three-level abnormal response, realizing working condition dynamic adaptation, S3, fault injection type pre-judging and maintenance, namely injecting a fault scene training pre-judging model into the digital twin body of S1, realizing fault positioning and root tracing through combining a five-dimensional knowledge map, S4, performing iterative evolution through knowledge driving, namely converting operation and maintenance related data of the previous three steps into knowledge map entities and relations, realizing cross-factory area experience multiplexing through transfer learning, and feeding the optimization strategy back to a pre-sequence step, and forming a complete operation and maintenance closed loop.

Inventors

YANG YUFENG
LIU YONGTAO
ZHANG JUNPENG
WAN JINBO
PAN SHIJUN
ZHANG JIE
Han Huiheng

Assignees

大唐国际发电股份有限公司张家口发电分公司

Dates

Publication Date: 20260505
Application Date: 20260206

Claims (10)

1. The pneumatic ash conveying operation and maintenance control method is characterized by comprising the following steps of: S1, intelligent perception and digital modeling, namely acquiring operation data and material characteristic data of a pneumatic ash conveying core component through distributed perception nodes, preprocessing through edge computing nodes, and constructing a digital twin body virtual mapping and standardized operation and maintenance database by combining federal learning to provide data support for subsequent steps; S2, multi-agent cooperative regulation and control, namely generating an optimal self-adaptive regulation and control strategy through real-time interaction of the multi-agent based on the virtual mapping and the database of the S1, and synchronously executing three-level abnormal response to realize dynamic adaptation of working conditions; S3, fault injection type prejudging maintenance, namely injecting a fault scene training prejudging model into the digital twin body in the S1, and realizing fault positioning and root tracing by combining a five-dimensional knowledge graph, so as to ensure stable operation; S4, knowledge-driven iterative evolution, namely converting the operation and maintenance related data of the first three steps into knowledge graph entities and relations, realizing cross-factory experience multiplexing through transfer learning, and feeding back an optimization strategy to the preamble step to form a complete operation and maintenance closed loop.
2. A pneumatic ash handling operation and maintenance control method according to claim 1, wherein the operation data includes at least one of pressure, wind speed, material level, temperature, vibration value, and the material characteristic data includes at least one of material moisture content and particle size distribution.
3. The pneumatic ash conveying operation and maintenance control method according to claim 1, wherein the preprocessing comprises the steps of cleaning, exception removing and feature extracting of original data, a layered processing mode of local early warning of exception data and cloud aggregation of key data is achieved, and the three-level exception response comprises the automatic regulation and control of edge-end agents in slight exception, the cooperative intervention of multiple agents in moderate exception, the starting of an emergency shutdown plan in severe exception and the pushing of early warning.
4. The pneumatic ash handling operation and maintenance control method according to claim 1, wherein the five-dimensional knowledge graph specifically comprises an equipment dimension, a parameter dimension, a material dimension, an environment dimension and a maintenance dimension; the equipment dimension comprises the model number, the operation time length and the historical fault record of the pneumatic ash conveying core component; The parameter dimension comprises pressure, wind speed, material level, temperature and vibration value operation parameters; the material dimension comprises material water content and particle size distribution material characteristic data; the environmental dimension includes an ambient temperature, humidity, and atmospheric pressure; the maintenance dimension comprises a fault handling scheme, an overhaul record and an operation and maintenance strategy.
5. The pneumatic ash conveying operation and maintenance control method according to claim 1, wherein the transfer learning realization of cross-factory experience multiplexing comprises the adaptation of factory individualized data fine adjustment parameters of a target factory by taking a factory basic model of a same industry standard as a source domain model.
6. The pneumatic ash conveying operation and maintenance control system is characterized by comprising a distributed sensing module, an edge calculation module, a digital twin module, a multi-agent regulation and control module, a fault pre-judgment module and a knowledge iteration module, wherein the modules are sequentially subjected to data interaction, and the output end of the knowledge iteration module is connected with the input ends of the multi-agent regulation and control module and the fault pre-judgment module to form a closed loop.
7. A pneumatic ash handling operation and maintenance control system according to claim 6, wherein the distributed sensing module comprises a plurality of sensing nodes disposed in the feeding device, the air source device, the conveying pipeline and the separation discharging device, and the sensing nodes at least comprise one or more of a pressure sensor and a wind speed sensor.
8. The pneumatic ash conveying operation and maintenance control system according to claim 6, wherein the multi-agent regulation and control module comprises a feeding agent, an air source agent, a pipeline air supplementing agent, a discharging agent and a reinforcement learning decision unit, and the agents interact in real time through an edge end communication module.
9. The pneumatic ash conveying operation and maintenance control system according to claim 6, wherein the fault pre-judging module comprises a fault scene injection unit, a model training unit and a tracing unit, and the five-dimensional knowledge graph is built in the tracing unit.
10. The pneumatic ash conveying operation and maintenance control system according to claim 6, wherein the knowledge iteration module comprises a knowledge map updating unit, a transfer learning unit and a strategy pushing unit, and the strategy pushing unit is used for pushing an optimization strategy to the preamble corresponding module.

Description

Pneumatic ash conveying system and operation and maintenance control method Technical Field The invention relates to the technical field of pneumatic ash conveying, in particular to a pneumatic ash conveying system and an operation and maintenance control method. Background The pneumatic ash conveying system in the industrial fields of thermal power, building materials and the like is core equipment for conveying powder materials, the thermal power field is used for conveying fly ash generated by boiler combustion, the building material field is used for bearing the transportation task of cement raw materials and clinker powder, the running environment is mainly high-temperature and dusty working conditions, the equipment load is frequent along with fluctuation of production line productivity, the system is extremely easily influenced by the water content of the fly ash and the environmental humidity change of a factory, the water content is too high, the material adhesion pipeline is easily caused, the environmental humidity is too high, the material agglomeration is aggravated, the conveying efficiency is directly disturbed, and even the equipment is blocked and stopped. In the traditional mode, when symptoms such as sudden pressure rise and abnormal vibration occur in the system, operation and maintenance personnel are required to disassemble and examine on site, so that the operation and maintenance personnel not only consume hours or even days, but also frequently cause repeated faults due to the fact that equipment, materials, environments and other factors are not easy to clean, and the unplanned outage rate is high, and therefore the pneumatic ash conveying system and the operation and maintenance control method are required to be provided. Disclosure of Invention The invention aims to solve the defects in the prior art, and provides a pneumatic ash conveying system and an operation and maintenance control method. In order to achieve the above purpose, the present invention adopts the following technical scheme: a pneumatic ash conveying operation and maintenance control method comprises the following steps: S1, intelligent perception and digital modeling, namely acquiring operation data and material characteristic data of a pneumatic ash conveying core component through distributed perception nodes, preprocessing through edge computing nodes, and constructing a digital twin body virtual mapping and standardized operation and maintenance database by combining federal learning to provide data support for subsequent steps; S2, multi-agent cooperative regulation and control, namely generating an optimal self-adaptive regulation and control strategy through real-time interaction of the multi-agent based on the virtual mapping and the database of the S1, and synchronously executing three-level abnormal response to realize dynamic adaptation of working conditions; S3, fault injection type prejudging maintenance, namely injecting a fault scene training prejudging model into the digital twin body in the S1, and realizing fault positioning and root tracing by combining a five-dimensional knowledge graph, so as to ensure stable operation; S4, knowledge-driven iterative evolution, namely converting the operation and maintenance related data of the first three steps into knowledge graph entities and relations, realizing cross-factory experience multiplexing through transfer learning, and feeding back an optimization strategy to the preamble step to form a complete operation and maintenance closed loop. The technical scheme further comprises the following steps: specifically, the operation data comprise at least one of pressure, wind speed, material level, temperature and vibration value, and the material characteristic data comprise at least one of material water content and particle size distribution. Specifically, the preprocessing comprises the steps of cleaning, exception removing and feature extraction of original data, and a layered processing mode of local early warning of exception data and cloud aggregation of key data is realized, and the three-level exception response comprises the autonomous regulation and control of edge end agents in slight exception, the cooperative intervention of multiple agents in moderate exception, and the starting of an emergency shutdown plan and the pushing of early warning in serious exception. Specifically, the five-dimensional knowledge graph specifically comprises an equipment dimension, a parameter dimension, a material dimension, an environment dimension and a maintenance dimension; the equipment dimension comprises the model number, the operation time length and the historical fault record of the pneumatic ash conveying core component; The parameter dimension comprises pressure, wind speed, material level, temperature and vibration value operation parameters; the material dimension comprises material water content and particle size distribution material characteris