一种树脂生产控制系统的异常报警分析方法及系统

By constructing a standard reference curve and calculating the local transient response energy and coupling deviation verification coefficient, the problem of alarm proliferation and missed alarms in the resin production control system was solved, and adaptive monitoring and multivariate coupling verification were realized, thereby improving the safety and reliability of the production process.

CN122116577BActive Publication Date: 2026-07-17LUOYANG REFINING & CHEM AOYOU CHEM CO LTD +1

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
LUOYANG REFINING & CHEM AOYOU CHEM CO LTD
Filing Date
2026-04-29
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

Existing resin production control systems suffer from both excessive alarms and missed alarms during non-steady-state phases, failing to distinguish between normal drastic changes and abnormal runaway trends. Single-variable monitoring leads to misjudgments of faults such as sensor drift, increasing safety risks in the production process.

Method used

By acquiring historical golden batch data of the resin production process, a standard reference curve is constructed, and the local transient response energy and coupling deviation verification coefficient of real-time production data are calculated. Combining the first and second derivatives, a comprehensive anomaly score is constructed to achieve adaptive monitoring and multivariate coupling verification.

Benefits of technology

It achieves adaptive monitoring under all operating conditions, improves the safety of the resin production process, prevents missed and false alarms, accurately identifies hidden faults such as sensor drift or minor leaks, and simplifies the judgment of operators.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122116577B_ABST
    Figure CN122116577B_ABST
Patent Text Reader

Abstract

本发明涉及化工过程控制技术领域,具体涉及一种树脂生产控制系统的异常报警分析方法及系统,方法包括:获取树脂生产过程的历史黄金批次数据,通过动态时间规整算法将时间轴对齐并构建标准参考曲线;采集实时生产数据,计算局部瞬态响应能量,表征当前工艺状态的平稳程度;计算实时生产数据与标准参考曲线之间的原始偏差,利用局部瞬态响应能量对原始偏差进行加权放大处理,得到加权动态偏差指数;构建耦合校验模型,计算耦合偏离度校验系数,结合加权动态偏差指数得到综合异常得分,当得分超过阈值曲线时触发报警。本发明通过能量加权与多变量耦合分析,解决非稳态阶段的误报与漏报问题,提高了树脂生产过程的安全性。
Need to check novelty before this filing date? Find Prior Art