CN-122009932-A - Floor recognition method and device

CN122009932ACN 122009932 ACN122009932 ACN 122009932ACN-122009932-A

Abstract

The application discloses a floor identification method and a floor identification device, and relates to the technical field of terminals. The method comprises the steps of obtaining elevator session data of a first building, which are sent anonymously by a plurality of terminals, wherein the elevator session data comprise starting elevator hall access point AP scanning information, ending elevator hall AP scanning information and elevator car AP scanning information sequences, performing space-time joint clustering on the starting elevator hall AP scanning information and the ending elevator hall AP scanning information to generate a plurality of elevator hall AP class clusters representing different floors, each elevator hall AP class cluster corresponds to one floor AP anchor point, each floor AP anchor point comprises AP characteristic information of the elevator hall, converting a scanning time sequence corresponding to each elevator car AP scanning information sequence into a spatial sequence among the floor AP anchor points based on an attention mechanism of AP signal intensity to obtain a floor anchor point time sequence, and determining the floor AP anchor point topology of the first building based on all the floor anchor point time sequences.

Inventors

LAI WENLONG

Assignees

维沃移动通信有限公司

Dates

Publication Date: 20260512
Application Date: 20260319

Claims (10)

1. A floor recognition method, comprising: Acquiring elevator session data of a first building, which are anonymously transmitted by a plurality of terminals, wherein the elevator session data comprise AP scanning information of an access point of an elevator hall at a starting point, AP scanning information of the elevator hall at a finishing point and an AP scanning information sequence of a car; Performing space-time joint clustering on the starting elevator hall AP scanning information and the ending elevator hall AP scanning information to generate a plurality of elevator hall AP class clusters representing different floors, wherein each elevator hall AP class cluster corresponds to a floor AP anchor point, and the floor AP anchor point comprises AP characteristic information of a floor elevator hall; Converting the scanning time sequence corresponding to each car AP scanning information sequence into a space sequence between floor AP anchor points based on a signal strength attention mechanism to obtain a floor anchor point time sequence; And determining the floor AP anchor point topology of the first building based on all the floor anchor point time sequences.
2. The method of claim 1, wherein the spatiotemporal joint clustering of the starting hoistway AP scan information and the ending hoistway AP scan information to generate a plurality of hoistway AP class clusters representing different floors comprises: Based on signal intensity, performing space-time joint clustering on the starting elevator hall AP scanning information and the ending elevator hall AP scanning information to generate a plurality of initial class clusters; For any two initial class clusters, determining the reference riding time length between the two initial class clusters according to at least one pair of starting point-end point elevator hall AP scanning information corresponding to the two initial class clusters; And based on the reference riding time length, carrying out time verification and screening on the elevator hall AP scanning information in the two initial class clusters to obtain the plurality of elevator hall AP class clusters.
3. The method of claim 1, wherein the AP signal strength based attention mechanism converts the scanning time sequence corresponding to each car AP scanning information sequence to a spatial sequence between floor AP anchor points, resulting in a floor anchor point time sequence, comprising: Taking the car AP scanning information in each car AP scanning information sequence as an inquiry item, and calculating the attention value between each car AP scanning information and the plurality of floor AP anchor points; Taking the floor AP anchor point with the highest attention value in the plurality of floor AP anchor points as the floor AP anchor point corresponding to the AP scanning information of each car; And sequencing the floor AP anchor points corresponding to each car AP scanning information in each car AP scanning information sequence according to the time sequence corresponding to each car AP scanning information sequence to obtain a floor anchor point time sequence.
4. The method of claim 1, wherein the determining the floor AP anchor topology of the first building based on the all floor anchor time series comprises: constructing a directed graph which takes a floor AP anchor point as a vertex and takes the floor anchor point time sequence as a side based on all floor anchor point time sequences; And performing topological sorting on the directed graph to obtain the floor AP anchor point topology of the first building.
5. The method of claim 1, wherein the elevator session data further comprises at least one weightless event message, at least one overweight event message; the method further comprises the steps of: Calculating the displacement layer number corresponding to the elevator session data according to the starting elevator hall AP scanning information and the ending elevator hall AP scanning information in the elevator session data and the floor AP anchor point topology of the first building; Determining a motion time sequence characteristic of an elevator operation process according to elevator event time corresponding to elevator session data, wherein the elevator event time comprises elevator starting time, elevator ending time, at least one weightless event time and at least one overweight event time; constructing an elevator motion model based on the elevator operation time sequence characteristics and the displacement layer number; the model parameters of the elevator motion model are used for indicating the corresponding relation between the displacement layer number and the elevator riding time in the elevator riding process.
6. The method of claim 5, wherein constructing an elevator motion model based on the elevator travel time series characteristics and the number of displacement floors comprises: And fitting physical parameters of an elevator kinematics equation based on the motion time sequence characteristics and the displacement layer number to obtain a parameterized elevator motion model.
7. A floor recognition method, comprising: After detecting that a first elevator session in a first building is ended, determining a terminal floor corresponding to the first elevator session based on first terminal elevator hall AP scanning information in first session data and floor AP anchor point topology of the first building, wherein the first session data is elevator session data of the first elevator session; the floor AP anchor point topology is constructed based on elevator session data of the first building, which are anonymously reported by a plurality of terminals.
8. The method of claim 7, wherein the first session data further comprises first landing hoistway AP scan information, weight loss event information and weight overload event information during the first elevator session; the method further comprises the steps of: calculating a starting floor corresponding to the first elevator session according to the first elevator hall AP scanning information and the floor AP anchor point topology; Calculating the displacement layer number of the first elevator session through an elevator motion model according to the first elevator event time corresponding to the starting floor and the first session data, wherein the first elevator event time comprises elevator starting time, elevator ending time, at least one weightless event time and at least one overweight event time; Calculating a first floor based on the origin floor and the displacement floor number; The destination floor is calibrated based on the first floor.
9. A floor recognition device, characterized by comprising: The receiving module is used for acquiring elevator session data of a first building, which are anonymously transmitted by a plurality of terminals, wherein the elevator session data comprise AP scanning information of an access point of an elevator hall at a starting point, AP scanning information of the elevator hall at a terminal point and an AP scanning information sequence of a car; A processing module for: Carrying out space-time joint clustering on the starting elevator hall AP scanning information and the finishing elevator hall AP scanning information received by the receiving module to generate a plurality of elevator hall AP clusters representing different floors, wherein each elevator hall AP cluster corresponds to a floor AP anchor point, and the floor AP anchor point comprises AP characteristic information of a floor elevator hall; Converting the scanning time sequence corresponding to each car AP scanning information sequence into a space sequence between floor AP anchor points based on a signal strength attention mechanism to obtain a floor anchor point time sequence; And determining the floor AP anchor point topology of the first building based on all the floor anchor point time sequences.
10. A floor recognition device, characterized by comprising: The processing module is used for determining a terminal floor corresponding to a first elevator session based on first terminal elevator hall AP scanning information in first session data and floor AP anchor point topology of the first building after detecting that the first elevator session in the first building is ended, wherein the first session data is elevator session data of the first elevator session; the floor AP anchor point topology is constructed based on elevator session data of the first building, which are anonymously reported by a plurality of terminals.

Description

Floor recognition method and device Technical Field The application belongs to the technical field of terminals, and particularly relates to a floor identification method and a floor identification device. Background With the acceleration of the urban process and the densification of high-rise buildings, the position sensing service of the terminal is extending from the traditional two-dimensional plane to the three-dimensional vertical space. The method has the advantages of realizing accurate positioning of the floor level in the building, and having important significance for optimizing the performance of a communication network, improving indoor navigation, facilitating travel experience and other applications. Currently, electronic devices may perform floor identification based on wireless access point fingerprints, such as Wi-Fi and cellular network signals. Specifically, the floor is determined by matching the AP signal characteristics scanned by the terminal with a pre-stored AP fingerprint database. However, the floor identification method has the problems that on one hand, the manual offline acquisition of AP fingerprint data is relied on, so that a fingerprint database is high in maintenance cost and sparse in sample, the increase and decrease of AP equipment and signal fluctuation are difficult to cope with, and on the other hand, when the electronic equipment scans AP signals, space instability is easy to occur due to signal shielding in a complex indoor environment, and positioning deviation is caused. As such, the accuracy and robustness of floor recognition in the related art are poor. Disclosure of Invention The embodiment of the application aims to provide a floor identification method and a floor identification device, which can improve the accuracy and the robustness of floor identification. The embodiment of the application provides a floor identification method, which comprises the steps of obtaining elevator session data of a first floor, which are anonymously sent by a plurality of terminals, wherein the elevator session data comprise starting elevator hall access point AP scanning information and ending elevator hall AP scanning information, a car AP scanning information sequence, carrying out space-time joint clustering on the starting elevator hall AP scanning information and the ending elevator hall AP scanning information to generate a plurality of elevator hall AP clusters representing different floors, each elevator hall AP cluster corresponds to one floor AP anchor point, the floor AP anchor points comprise AP characteristic information of the elevator hall of the floor, converting a scanning time sequence corresponding to each car AP scanning information sequence into a spatial sequence among the floor AP anchor points based on a signal intensity attention mechanism to obtain a floor anchor point time sequence, and determining the floor AP anchor point topology of the first floor based on all the floor anchor point time sequences. The embodiment of the application provides a floor identification device, which comprises a receiving module and a processing module, wherein the receiving module is used for acquiring elevator session data of a first floor, the elevator session data comprise starting elevator hall access point AP scanning information and destination elevator hall AP scanning information, the elevator car AP scanning information is a car AP scanning information sequence, the processing module is used for carrying out space-time joint clustering on the starting elevator hall AP scanning information and the destination elevator hall AP scanning information received by the receiving module to generate a plurality of elevator hall AP clusters representing different floors, each elevator hall AP cluster corresponds to a floor AP anchor point, the floor AP anchor point comprises AP characteristic information of a floor elevator hall, a signal intensity-based attention mechanism is used for converting a scanning time sequence corresponding to each car AP scanning information sequence into a spatial sequence between floor anchor points to obtain a floor anchor point time sequence, and the floor anchor point AP topology of the first floor is determined based on all floor anchor point time sequences. In a third aspect, an embodiment of the present application provides a floor identifying method, where after a first elevator session in a first building is detected to end, determining a destination floor corresponding to the first elevator session based on first destination elevator hall AP scan information in first session data and a floor AP anchor topology of the first building, where the first session data is elevator session data of the first elevator session, and the floor AP anchor topology is constructed based on elevator session data of the first building that is anonymously reported by a plurality of terminals. In a fourth aspect, an embodiment of the present applica