Vehicle Connection Time Prediction With Confidence-Based Polling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Managing a large number of vehicles to function as virtual power plants generates excessive data traffic, leading to increased communication costs for servers that predict and coordinate their connection times to the power grid.
Innovation Solution
A server system that acquires location information from vehicles not connected to the power grid, uses a prediction model to estimate connection times, and adjusts communication frequency based on the confidence of the prediction, reducing data traffic and communication costs by lowering the frequency when confidence is high and increasing it when confidence is low.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the server frequently acquires data from all vehicles to accurately predict connection times, then prediction accuracy is improved, but communication costs increase
Solution Approach 1:
The server applies different data acquisition frequencies to different vehicles based on their individual prediction confidence levels. Vehicles with low prediction confidence (high uncertainty) receive frequent data acquisition, while vehicles with high prediction confidence receive infrequent data acquisition. This localized differentiation resolves the contradiction by optimizing communication resources according to actual prediction needs rather than applying a uniform frequency to all vehicles.
Solution Approach 2:
The data acquisition frequency is dynamically adjusted based on the calculated prediction confidence for each vehicle. As prediction confidence changes over time (based on new location data and updated prediction models), the acquisition frequency automatically adapts. This dynamic adjustment allows the system to maintain high prediction accuracy when needed while reducing communication costs when prediction is already confident.
2Reliability
If the server maintains high data acquisition frequency for all vehicles, then connection time prediction reliability is improved, but operational expenses increase
Solution Approach 1:
The server changes the parameter of data acquisition frequency based on the prediction confidence parameter for each vehicle. By calculating confidence metrics from prediction models and using these to adjust acquisition frequencies, the system maintains prediction reliability for vehicles that need it while improving operational efficiency by reducing unnecessary data acquisition from vehicles with stable, predictable patterns.
3Productivity
If the server reduces data acquisition frequency to lower communication costs, then operational expenses are reduced, but prediction accuracy deteriorates
Solution Approach 1:
Instead of uniformly reducing data acquisition frequency across all vehicles, the server applies frequency reduction selectively to vehicles with high prediction confidence. Vehicles with low prediction confidence continue to receive frequent data acquisition to maintain accuracy. This localized approach allows the system to improve operational efficiency through reduced communication while preserving prediction accuracy where it matters most.
Solution Approach 2:
The system uses feedback from prediction confidence calculations to adjust data acquisition frequencies. The prediction model continuously evaluates confidence based on acquired location data, and this confidence feedback drives the decision-making process for frequency adjustment. This closed-loop feedback mechanism ensures that prediction accuracy is maintained through data acquisition only when and where confidence levels indicate it is necessary.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A vehicle management server (3) includes: a communication device (33) that acquires data including location information of a target vehicle through communication, the target vehicle being a vehicle not connected to the power grid out of vehicles (1); and a processor (311) that sets a frequency of communication at which the data is acquired from the target vehicle. The processor (311) acquires the data from the target vehicle. The processor (311) predicts connection time from the acquired data according to a time prediction model (5) and calculates confidence of the predicted connection time, the connection time being time when the target vehicle becomes connectable to the power grid, and the time prediction model (5) being a model that predicts the connection time based on the data from the target vehicle. The processor (311) sets the frequency of communication according to the confidence.