Elevator Call Allocation via Passenger Batch Distributions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing elevator allocation methods, such as continuous and immediate allocation, often fail to accurately determine the number of passengers waiting behind a call, leading to inefficient allocation decisions, especially in situations where the estimated number exceeds available elevator space.
Innovation Solution
Constructing passenger batch size distributions based on actual journeys between floors, including origin, destination, and number of passengers, to estimate the number of passengers waiting behind a call and allocate elevators accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If immediate allocation is used, then allocation decision speed is improved, but accuracy of passenger number estimation deteriorates
Solution Approach 1:
The system performs preliminary actions by collecting and storing passenger journey data (origin, destination, batch size, time) before allocation decisions are needed. This historical data is pre-processed to create probability distributions, enabling accurate real-time estimation without delaying the immediate allocation decision.
Solution Approach 2:
The system implements feedback by using actual passenger batch size data from completed journeys to continuously refine and update the probability distributions. This feedback loop improves the accuracy of passenger number estimates over time while maintaining the speed of immediate allocation decisions.
2Adaptability or versatility
If continuous allocation is used, then flexibility in allocation decisions is improved, but complexity of passenger number estimation deteriorates
Solution Approach 1:
The system transforms the complex estimation problem into a more manageable form by changing parameters from direct counting to probability-based estimation using pre-computed distributions. This allows flexible re-allocation decisions while simplifying the estimation complexity through statistical modeling.
Solution Approach 2:
Instead of directly tracking and estimating individual passenger queues in real-time (complex), the system creates simplified copies in the form of probability distributions that represent aggregate passenger batch sizes. These distributions can be quickly queried to support flexible allocation decisions without complex real-time estimation.
3Device complexity
If historical passenger arrival intensity is used, then estimation process is simplified, but accuracy of passenger number estimation deteriorates
Solution Approach 1:
The system segments the estimation approach by separating the simplicity of using historical data from the accuracy requirements. It divides passenger journeys into discrete batches with recorded actual sizes, creating segmented data points that form the basis of accurate probability distributions while maintaining a simple estimation process.
Solution Approach 2:
The system transitions from one-dimensional historical arrival intensity data to a two-dimensional approach by incorporating actual batch sizes as a new dimension. This adds the dimension of batch size distribution to the estimation process, significantly improving accuracy while keeping the process simple through pre-computed probability tables.
Data Source
Figure 1
Figure 2~3
AI summary
According to an example embodiment there is provided a method for allocating an elevator in an elevator system. The method comprises constructing passenger batch size distributions (100) for each pair of floors in a building based on passenger batch journeys, each passenger batch journey defining at least the origin and destination floor of the journey, the number of passengers relating to the journey and the time of the journey; receiving a call for an elevator (102); estimating the number of passengers waiting behind the call (104) based on the passenger batch size distributions; and allocating the call to an elevator being able to serve the estimated number of passengers (106).