Transformer Substation Load Prediction to Prevent Repeated Switching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for optimizing transformer substation operations result in repeated switching near intersection points, leading to increased energy consumption and reduced service life due to the need for plotting all possible operating models in advance, which is costly and time-consuming.
Innovation Solution
A method and apparatus that predict current and future loads of transformers, calculate costs including losses and operation costs, and optimize switch operations to minimize total costs by using formulas to determine optimal transformer and switch usage over time periods, preventing unnecessary switching and improving operational accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If the prior art method of plotting curves of all possible operating models is used to determine operating modes, then the transformer operation can be optimized, but repeated switching occurs near intersection points which increases energy consumption and reduces service life
Solution Approach 1:
The patent applies preliminary action by predicting future loads in advance before making switching decisions. The optimization method forecasts loads for multiple time periods ahead and determines switching strategies based on these predictions, rather than reacting to current conditions alone. This allows the system to anticipate intersection points and avoid unnecessary switching operations, thereby reducing energy consumption and extending transformer service life.
Solution Approach 2:
The patent implements dynamics by transitioning from static curve plotting to dynamic load prediction. Instead of using fixed pre-plotted curves, the system dynamically forecasts future loads based on historical data and current conditions, allowing the optimization strategy to adapt to changing conditions. This dynamic approach prevents repeated switching near intersection points by understanding future load trends.
2Productivity
If the prior art method plots curves of all possible operating models in advance, then operating modes can be determined, but this requires significant cost and time resources
Solution Approach 1:
The patent applies the taking out principle by extracting only the necessary information needed for optimization rather than plotting all possible operating curves. Instead of comprehensively analyzing every possible operating model in advance, the system selectively predicts future loads and identifies relevant switching opportunities, significantly reducing the time and computational resources required while maintaining optimization effectiveness.
Solution Approach 2:
The patent uses copying by creating simplified predictive models that replicate the essential behavior of complex operating scenarios without requiring full detailed analysis. The load prediction mechanism uses historical data patterns to generate future load profiles, which are then used for optimization decisions, avoiding the need to plot and analyze all possible operating curves in detail.
3Adaptability or versatility
If frequent switching occurs near intersection points, then operating mode adjustments can be made, but this reduces service life of transformers and increases power consumption
Solution Approach 1:
The patent implements feedback by continuously monitoring actual load conditions and comparing them with predicted loads, then using this information to refine switching decisions. The system learns from past switching outcomes and load patterns, adjusting its prediction and switching strategy to minimize unnecessary operations. This feedback mechanism maintains operational flexibility while reducing power consumption by avoiding redundant switching near intersection points.
Data Source
AI summary
Various embodiments of the teachings herein include a method for optimizing operation of a transformer substation with transformers connected or disconnected by switches. The method may include: acquiring a current load of each transformer, and predicting loads for each in various operating modes in some time periods based on current loads; calculating transformer costs for each time period based on predicted loads including transformer losses and operation costs; calculating switch costs based on purchase cost and service life for a switch corresponding to each transformer; and calculating a total cost in a total time period using the transformer costs and the switch costs, optimizing the total cost to obtain optimization parameters, and operating the transformer substation according to the optimization parameters, the total time period consisting of the plurality of time periods.