Process Trend Display With Secondary Alarm Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing visualization systems for industrial technical systems fail to reliably display trend curves and forecasts due to smoothing effects from averaging or median value formation, leading to missed secondary alarms when the display period is too long, and distortions in future forecasts.
Innovation Solution
The visualization system determines the minimum and maximum values of process data points during the acquisition period and displays secondary alarms when these values exceed predetermined thresholds, ensuring that secondary alarms are not missed due to averaging or median value formation, and provides improved forecasting by displaying predicted alarms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Duration of action of stationary object
If the display period is extended to show longer-term trends, then the overview of process development is improved, but secondary alarms may be missed due to smoothing effects from averaging or median value formation
Solution Approach 1:
The patent segments the process data representation into multiple layers: the main trend curve uses smoothed values (mean or median) for overview, while a separate parallel representation shows raw individual measured values. This segmentation allows the smoothed curve to provide trend overview while the raw values layer detects secondary alarms without being affected by smoothing, thus resolving the contradiction between extended display period and alarm detection reliability.
2Ease of operation
If averaging or median value formation is used to smooth the trend curve, then the overview of process development is improved, but secondary alarms are missed and forecasts are distorted
Solution Approach 1:
The patent introduces an intermediary representation layer that displays raw individual measured values in parallel with the smoothed trend curve. This intermediary layer acts as a mediator that preserves alarm information lost in the smoothing process, allowing operators to evaluate trends easily via the smoothed curve while simultaneously detecting secondary alarms through the raw values representation, thus preventing information loss.
3Productivity
If forecasts are generated based on smoothed trend curves, then future process development prediction is provided, but forecast accuracy is significantly distorted by smoothing
Solution Approach 1:
The patent segments forecast information into two distinct representations: forecasts based on smoothed trend curves for general process development prediction, and separate forecasts based on raw individual measured values for accurate alarm prediction. This segmentation maintains productivity by providing forecasting capability while improving measurement precision by using appropriate data sources for different forecast purposes, preventing accuracy distortion.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
An industrial technical plant (1) is controlled and monitored by a process control system (2). The process control system (2) collects process data (x, x*, Pi) at collection times and archives the collected process data (x, x*, Pi) as a history. On the basis of the selection of a process datum (x, x*, Pi) and the specification of an associated display period by an operator (4), a visualisation system (5) requests the history of the selected process datum (x, x*, Pi) for that display period from the process control system (2) and outputs the associated profile over time as a graphic (8) to the operator (4). The graphic (8) comprises a number of process data points (9). The visualisation system (5) determines the particular process data point (9) as the mean or median of the values of the selected process datum (x, x*, Pi) in a collection period that is characteristic of the particular process data point (9). For the particular process data point (9), it also determines the minimum (Min) and/or the maximum (Max) of the values of the selected process datum (x, x*, Pi) during the collection period. Whenever the minimum (Min) determined for the particular process data point (9) is below a predefined minimum value (MIN) and/or the maximum (Max) determined exceeds a predefined maximum value (MAX), the visualisation system (5) indicates, together with the particular process data point (9), a secondary alarm (A'), a secondary alarm (A') being an alarm that is not set by the process control system (2), but by the visualisation system (5).