Pump Efficiency Control Using ML and VFD Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for controlling motor-driven equipment and systems, particularly those connected to Variable Frequency Drives (VFDs, have not seen significant improvements in efficiency and precision over the past 150 years, leading to suboptimal energy consumption and mechanical strain.
Innovation Solution
Implementing a machine learning model trained with efficiency and operational characteristic data from characteristic curves to predict real-time efficiency, optimizing the operation of motor-driven systems by turning devices on or off based on determined efficient scenarios, using a combination of mathematical techniques and artificial intelligence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional control methods are used to regulate motor-driven equipment, then the system achieves desired results with precision, but energy consumption is suboptimal and mechanical strain increases
Solution Approach 1:
The system continuously monitors actual operating conditions (flow rate, head, power consumption) and compares them with optimal values from performance curves. The controller adjusts motor speed via VFD to maintain operation at or near the Best Efficiency Point, creating a closed-loop feedback system that optimizes energy consumption while maintaining control precision.
Solution Approach 2:
The system dynamically adjusts motor speed based on real-time operating conditions rather than using fixed speed settings. By continuously varying the motor speed to match optimal points on performance curves, the system adapts to changing load requirements and maintains peak efficiency across different operating scenarios.
2Productivity
If conventional control methods are used to regulate motor-driven equipment, then the system achieves desired results, but mechanical strain increases leading to reduced equipment life
Solution Approach 1:
The feedback mechanism monitors operating conditions and prevents the system from operating in high-strain regions by adjusting speed to stay within optimal parameters. This continuous monitoring and adjustment prevents excessive mechanical stress on motor and pump components, extending equipment life while maintaining productivity.
Solution Approach 2:
The system uses pre-stored performance curve data that identifies optimal operating points before actual operation occurs. By having this optimization data ready in advance and using it to guide control decisions, the system proactively prevents mechanical strain before it occurs, rather than reacting to damage after it happens.
3Use of energy by moving object
If machine learning model with complete system-efficiency dataset is implemented, then energy consumption is reduced, but device complexity increases
Solution Approach 1:
The system performs the computationally intensive data extraction, extrapolation, and model training in advance, before actual operational optimization begins. By pre-processing performance curve data and creating the efficiency model offline, the system avoids real-time computational complexity while still achieving sophisticated energy optimization during operation.
Solution Approach 2:
The patent replaces complex real-time mathematical calculations and physical measurements with a pre-trained machine learning model that makes predictions based on input operating conditions. This substitution of a complex computational system with a simpler predictive model reduces real-time processing requirements while maintaining optimization accuracy.
Data Source
AI summary
Embodiments provide functionality to control real-world mechanical systems through the creation and deployment of machine learning models. An embodiment creates the machine learning model by extracting (i) an indication of efficiency and (ii) values of operational characteristics of one or more devices from one or more characteristic curves. Each characteristic curve corresponds to a respective device of one or more devices, in a mechanical system, functioning at a given speed. A training data set is created by determining efficiency and values of the operational characteristics for the mechanical system functioning with multiple combinations of the one or more devices operating at each of a plurality of speeds using the extracted indication of efficiency and extracted values of the operational characteristics. In turn, the machine learning model is trained with the created training dataset. Training configures the machine learning model to predict efficiency of the mechanical system based on operating data.