EV Charging Station Current Prediction for Spoofed Data Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional methods for detecting erroneous data and data spoofing at electric vehicle charging stations are costly and complex, complicating their implementation and impacting the power drawn from the electrical utility.
Innovation Solution
A controller and method that utilize multiple trained data models to predict the current drawn by the electric vehicle charging station, ignoring measurements from specific equipment to identify erroneous data without requiring additional sensors or equipment, thereby reducing costs and complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional methods for detecting erroneous data and data spoofing are implemented, then detection capability is improved, but device complexity and cost increase
Solution Approach 1:
The system uses its own existing measurements and data models to detect erroneous data, without requiring external detection equipment. The controller performs self-diagnosis by comparing predicted current values (generated from data models that ignore specific EVSE measurements) with actual measured current, allowing the system to identify which EVSE is generating erroneous data using only its own resources
Solution Approach 2:
The system creates virtual copies of measurement data by generating predicted current values through multiple data models, each ignoring a different EVSE's measurements. These predicted values serve as virtual representations that can be compared against actual measurements to detect anomalies without needing physical duplicate sensing equipment
2Reliability
If traditional methods for detecting erroneous data and data spoofing are implemented, then detection capability is improved, but manufacturing cost increases
Solution Approach 1:
The system uses its own existing measurements and data models to detect erroneous data, without requiring external detection equipment. The controller performs self-diagnosis by comparing predicted current values (generated from data models that ignore specific EVSE measurements) with actual measured current, allowing the system to identify which EVSE is generating erroneous data using only its own resources
Solution Approach 2:
The system creates virtual copies of measurement data by generating predicted current values through multiple data models, each ignoring a different EVSE's measurements. These predicted values serve as virtual representations that can be compared against actual measurements to detect anomalies without needing physical duplicate sensing equipment
3Measurement precision
If multiple data models are used to predict current values, then detection accuracy is improved, but processing complexity increases
Solution Approach 1:
The detection process is segmented into distinct steps: (1) Generate multiple predicted current values using different data models, each ignoring a specific EVSE's measurements; (2) Compare each predicted value with the actual measured current; (3) Identify the EVSE whose ignored measurement corresponds to the largest discrepancy. This segmentation makes the complex multi-model comparison manageable and systematic
Solution Approach 2:
The system generates more predicted current values than strictly necessary (using one data model per EVSE plus potentially additional models), providing excess detection coverage. This excessive action ensures robust detection capability while the systematic comparison process keeps processing complexity manageable
Data Source
Figure 1
Figure 2
Figure 3
AI summary
In one aspect, a controller (104) for detecting erroneous data generated at an electric vehicle charging station (102) (EVCS) is provided. The EVCS includes a plurality of electric vehicle supply equipment (106, 107, 108, 109) (EVSE) for charging electric vehicles. The controller is configured to store a plurality of data models (133, 134, 135, 136) that predict a current at a point of common coupling (112) (PCC) drawn by the EVCS from a utility (110), where each of the plurality of data models ignores measurements from a different one of the plurality of EVSEs, generate a plurality of predicted current values (138, 139, 140, 141), each generated using a different one of plurality of data models, measure an actual current value (132) at the PCC, calculate a plurality of difference values, each comprising a difference between one of the predicted current values and the actual current value, and determine whether the erroneous data is being generated based on the plurality of difference values.