Cooling Tower Monitoring with Kalman Filter Bias Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cooling towers face challenges in accurately monitoring and diagnosing issues due to sensor biases, which can lead to costly and detrimental corrections if not accurately measured, affecting the entire system's operation.
Innovation Solution
A simulation model using an extended Kalman filter and virtual controllers to correct sensor biases, estimate unmeasurable variables, and generate accurate predictions for monitoring and controlling cooling tower systems, incorporating historical data and real-time measurements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If sensor measurements are used to monitor cooling tower parameters, then real-time monitoring capability is improved, but measurement accuracy deteriorates due to sensor biases
Solution Approach 1:
A simulation model acts as an intermediary between sensor measurements and diagnostic decisions. The model integrates multiple sensor inputs and uses process knowledge to generate corrected parameter estimates, filtering out individual sensor biases while maintaining real-time monitoring capability.
Solution Approach 2:
The system uses feedback from the simulation model to continuously correct sensor readings. By comparing model predictions with actual measurements and adjusting for discrepancies, the system compensates for sensor biases in real-time without sacrificing monitoring speed.
2Difficulty of detecting and measuring
If sensor measurements with biases are used for diagnosis, then diagnostic capability is improved, but system reliability deteriorates due to incorrect diagnoses
Solution Approach 1:
The simulation model serves as a reliable intermediary that processes potentially biased sensor data through physics-based relationships and multiple data sources, producing more reliable diagnostic conclusions than any single sensor could provide.
Solution Approach 2:
The system transforms raw sensor measurements into corrected parameter estimates by applying calibration factors and corrections derived from the simulation model, changing the parameter representation from biased measurements to reliable estimates.
3Loss of energy
If corrections are made based on inaccurate measurements, then operational costs increase, but measurement accuracy remains unchanged
Solution Approach 1:
The simulation model provides feedback on the accuracy and reliability of measurements, allowing the system to avoid costly corrections when measurements are likely accurate and to make informed correction decisions only when necessary, reducing unnecessary operational costs.
Solution Approach 2:
The simulation model intermediates between raw measurements and correction actions, providing a reasoned assessment of whether corrections are truly needed based on model predictions and measurement consistency, thereby avoiding wasteful corrections.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A cooling tower simulation system may receive a measurement from a cooling tower sensor and generate a predicted output of a cooling tower system based on a model of the cooling tower system. The simulation system may generate an estimated output using an extended Kalman filter with the measurement and the predicted output as inputs, wherein the estimated output represents a characteristic of the cooling tower system.