CN-122009942-A - Escalator operation control method, escalator operation control system and computer equipment

CN122009942ACN 122009942 ACN122009942 ACN 122009942ACN-122009942-A

Abstract

The embodiment of the invention provides a escalator operation control method, a escalator operation control system and computer equipment, relates to the technical field of escalator operation control, and aims to solve the problems that traditional escalator regulation relies on manual experience and is difficult to cope with dynamic passenger flow changes so as to improve the intelligent level of regulation and operation safety, efficiency and energy efficiency. The method comprises the steps of obtaining operation source data of a controlled escalator, wherein the operation source data comprise historical operation data, historical passenger flow data and operation environment data, the historical operation data comprise historical operation speed control data of the controlled escalator, generating predicted passenger flow data of a future period according to the historical passenger flow data, generating controlled escalator operation risk prediction data of the future period based on the predicted passenger flow data and the historical operation speed control data of the controlled escalator, generating an escalator operation control strategy of the future period by utilizing the controlled escalator operation risk prediction data, and controlling the controlled escalator to operate in the future period according to the escalator operation control strategy.

Inventors

YANG XIAOXIA
JI SHUAI
XIA BIN
JIN SIBO
SHI PENGYUAN
Qi Sijun
Guo Kenhong
LI DANNA
HUANG SHOUXIANG
CAO XU
LI QIANG
MA HAO
YANG FAZHAN
WEI JINLI
CHENG SEN
PENG WENXIANG

Assignees

青岛理工大学
中铁建电气化局集团第三工程有限公司
中国铁建电气化局集团有限公司

Dates

Publication Date: 20260512
Application Date: 20260413

Claims (10)

1. An escalator operation control method is characterized by comprising the following steps: acquiring operation source data of a controlled escalator, wherein the operation source data comprises historical operation data, historical passenger flow data and operation environment data, and the historical operation data comprises historical operation speed control data of the controlled escalator; Generating predicted passenger flow data of a future period according to the historical passenger flow data; generating controlled escalator operation risk prediction data of the future period based on the predicted passenger flow data and the historical operation speed control data of the controlled escalator; generating an escalator operation control strategy of the future period by using the controlled escalator operation risk prediction data; And controlling the controlled escalator to operate in the future time period according to the escalator operation control strategy.
2. The method of claim 1, wherein generating controlled escalator operational risk prediction data for the future period based on the predicted passenger flow data and historical operational speed control data for the controlled escalator comprises: acquiring risk prediction key feature information, wherein the key features comprise passenger flow data, risk event data and hidden danger data; Constructing a risk prediction model based on different risk assessment tasks; Predicting risk indexes of the future period based on the risk prediction model by utilizing the key characteristic information, wherein the risk indexes comprise a passenger flow time index, a congestion degree index and a default risk index.
3. The method of claim 2, wherein predicting a risk indicator for the future period based on the risk prediction model using the key feature information comprises: calculating objective weights of each risk index in the future period by utilizing the historical passenger flow data and the risk event data; Calculating fuzzy subjective weight of each risk index in the future period based on the fuzzy judgment matrix; And fusing the objective weight and the fuzzy subjective weight by adopting a weighted average method to obtain the comprehensive weight of each risk index in the future period.
4. The method of claim 3, wherein predicting a risk indicator for the future period based on the risk prediction model using the key feature information further comprises: And updating the weight distribution of each risk index through a game theory model.
5. The method of claim 4, wherein predicting a risk indicator for the future period based on the risk prediction model using the key feature information further comprises: performing joint training on a plurality of risk assessment tasks; and fusing the output results of each risk assessment task in a weighted average fusion mode to obtain a comprehensive risk value.
6. The method of claim 5, wherein generating the escalator operation control strategy for the future time period using the controlled escalator operation risk prediction data comprises: constructing a state of the reinforcement learning intelligent agent by the predicted passenger flow data and the comprehensive risk value; taking a target running speed or state set of the controlled escalator as an action space of the reinforcement learning intelligent agent; and generating and outputting the operation control strategy of the controlled escalator through the reinforcement learning strategy of the reinforcement learning intelligent agent.
7. The method of claim 6, wherein generating the escalator operation control strategy for the future time period using the controlled escalator operation risk prediction data further comprises: training the reinforcement learning strategy using a multi-step scrolling strategy comprising: Defining the current reinforcement learning state of the reinforcement learning intelligent agent according to the predicted passenger flow data and the risk prediction data at the current moment; the reinforcement learning intelligent agent obtains instant rewards in the action space through designating the escalator speed; Updating the current reinforcement learning state.
8. The method of any of claims 1-5, wherein generating predicted passenger flow data for a future period from the historical passenger flow data comprises: converting the plane layout information of the space where the controlled escalator is located into two-dimensional grids or image data; Inputting the two-dimensional grid or the image data into a convolutional neural network, and extracting the spatial characteristics of the space where the controlled escalator is located; constructing a graph structure of a key area and related historical passenger flow data in a space where the controlled escalator is located; Based on the spatial features, a knowledge-guided graph convolution model is utilized to obtain the passenger flow volume or passenger flow density of each critical area within the future period.
9. An escalator operation control system, comprising: The system comprises a data acquisition module, a control module and a control module, wherein the data acquisition module is used for acquiring operation source data of a controlled escalator, the operation source data comprise historical operation data, historical passenger flow data and operation environment data, and the historical operation data comprise historical operation speed control data of the controlled escalator; the future passenger flow prediction module is used for generating predicted passenger flow data of a future period according to the historical passenger flow data; a future risk prediction module, configured to generate controlled escalator operation risk prediction data of the future period based on the predicted passenger flow data and the historical operation speed control data of the controlled escalator; The control strategy generation module is used for generating an escalator operation control strategy of the future period by utilizing the controlled escalator operation risk prediction data; And the operation control module is used for controlling the controlled escalator to operate in the future time period according to the escalator operation control strategy.
10. A computer device comprising at least one processor and at least one memory, the memory storing executable instructions for the processor, the processor configured to perform the escalator operation control method of any one of claims 1-8.

Description

Escalator operation control method, escalator operation control system and computer equipment Technical Field The invention belongs to the technical field of escalator operation control, in particular to the field of urban rail transit and passenger flow management, and particularly relates to an escalator operation control method, an escalator operation control system and computer equipment. Background The existing subway station is generally faced with high-density passenger flow in the period of the morning and evening peak and holiday, the conventional escalator dispatching which depends on manual experience or a fixed strategy is difficult to accurately cope with the rapid increase or mutation of the passenger flow and is easy to generate crowding and potential safety hazards, on the other hand, the pure improvement of the escalator speed brings additional energy consumption and bottleneck superposition risks, meanwhile, the lack of quantitative measurement on the illegal behaviors of passengers (such as forbidden and changed directions of the escalator) leads to the lack of a scientific comprehensive evaluation method for passenger flow time, crowding resistance and illegal cost, so that the passenger flow is predicted by combining a machine learning technology, and the dynamic self-adaptive control on the escalator speed is realized by utilizing a multi-index risk evaluation and reinforcement learning algorithm, thereby taking the safety and efficiency into consideration. Disclosure of Invention In order to at least partially solve the foregoing problems, the embodiments of the present invention provide an escalator operation control method, system and computer device that integrate machine learning passenger flow prediction, risk assessment and reinforcement learning decision. In a first aspect, an embodiment of the present invention provides an escalator operation control method, including: acquiring operation source data of a controlled escalator, wherein the operation source data comprises historical operation data, historical passenger flow data and operation environment data, and the historical operation data comprises historical operation speed control data of the controlled escalator; Generating predicted passenger flow data of a future period according to the historical passenger flow data; generating controlled escalator operation risk prediction data of the future period based on the predicted passenger flow data and the historical operation speed control data of the controlled escalator; generating an escalator operation control strategy of the future period by using the controlled escalator operation risk prediction data; And controlling the controlled escalator to operate in the future time period according to the escalator operation control strategy. Further, the generation of the controlled escalator operation risk prediction data of the future period based on the predicted passenger flow data and the historical operation speed control data of the controlled escalator comprises the steps of obtaining risk prediction key feature information, wherein the key feature information comprises passenger flow data, risk event data and hidden danger data, constructing a risk prediction model based on different risk assessment tasks, and predicting risk indexes of the future period based on the risk prediction model by utilizing the key feature information, wherein the risk indexes comprise passenger flow time indexes, crowdedness indexes and default risk indexes. Further, predicting the risk indexes of the future period based on the risk prediction model by utilizing the key characteristic information comprises the steps of calculating objective weights of each risk index in the future period by utilizing the historical passenger flow data and the risk event data, calculating fuzzy subjective weights of each risk index in the future period based on a fuzzy judgment matrix, and fusing the objective weights and the fuzzy subjective weights by adopting a weighted average method to obtain comprehensive weights of each risk index in the future period. Further, predicting risk indicators for the future time period based on the risk prediction model using the key feature information further includes updating a weight allocation for each of the risk indicators via a game theory model. Further, predicting the risk index of the future period based on the risk prediction model by utilizing the key characteristic information further comprises the steps of carrying out joint training on a plurality of risk assessment tasks, and fusing output results of each risk assessment task in a weighted average fusion mode to obtain a comprehensive risk value. Further, the escalator operation control strategy for generating the future time period by utilizing the controlled escalator operation risk prediction data comprises the steps of forming a state of an reinforcement learning intelligent agent by using the predict