Parking Space Forecasting Using Cell Grid Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for forecasting available parking spaces in urban areas are inefficient and costly, with low accuracy and high computing requirements, especially when relying on mobile phone data, due to the need to process extensive data sets and complex calculations.
Innovation Solution
The method selects significant cells in a cell grid based on statistical significance, using real-time and static data sets to reduce data processing needs, combining statistical and sensor data for accurate forecasting with reduced system costs, and employing machine learning for adjustments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If extensive mobile phone data sets are processed using complex calculations to forecast parking spaces, then forecasting accuracy is improved, but system costs and computing effort increase significantly
Solution Approach 1:
The patent divides the urban area into a cell grid structure and processes mobile phone data in discrete time intervals. By segmenting the data processing into manageable cells and time steps, the system achieves comprehensive coverage without requiring complex centralized processing of all data simultaneously, thus maintaining accuracy while reducing system complexity
Solution Approach 2:
The patent extracts only the essential features from mobile phone data that are relevant for parking space forecasting, such as presence/absence in cells and basic movement patterns. By taking out only the necessary data elements rather than processing complete raw datasets, the system maintains forecasting accuracy while significantly reducing computational complexity and system costs
2Measurement precision
If real-time mobile phone data from all cells is processed continuously, then prediction accuracy is improved, but computing resources and time consumption increase
Solution Approach 1:
The patent implements periodic data processing by evaluating mobile phone data at discrete time intervals rather than continuously. This periodic approach allows the system to maintain accurate real-time predictions while significantly improving computing efficiency, as resources are utilized in scheduled batches rather than continuous operation
Solution Approach 2:
The patent processes data for all cells in the grid but focuses computational effort on cells that are relevant to the user's current location and destination. By performing partial processing on a subset of cells rather than exhaustive processing of all possible data, the system maintains prediction accuracy for needed areas while improving overall computing efficiency
3Reliability
If comprehensive traffic data from multiple sources is integrated, then forecasting reliability is improved, but data processing complexity and system costs increase
Solution Approach 1:
The patent merges mobile phone data with static traffic information and sensor data from parking spaces into a unified forecasting model. By combining these different data sources through a integrated cell-based approach, the system improves forecasting reliability while managing processing complexity through the standardized grid structure that handles multiple data types uniformly
Data Source
Figure 1a~1b
Figure 2~3
Figure 4
AI summary
A method for predicting available parking spaces, comprising the following steps performed by a computing unit: - a traffic structure analysis in which significant cells are selected from cells of a cell grid based on a static dataset formed from traffic data from a first period; - a calculation in which a statistical parking situation is determined for at least one of the cells of the cell grid based on a real-time dataset formed from traffic data of the significant cells from a second period; and - in which a sensor parking situation is determined for at least one of the cells of the cell grid based on a sensor dataset formed from sensor data of the cells from a third period.and - a probability forecast in which, based on the statistical parking situation and the sensor parking situation, a forecast of available parking spaces is carried out for at least one of the cells of the cell grid.