Airplane passenger cabin management method and device

Abstract

The invention discloses a management method and device for an airplane passenger cabin. The management method comprises the steps that data of the airplane passenger cabin are collected through a passenger behavior sensing agent, an environment monitoring agent, an equipment state monitoring agent and a steward physiological state monitoring agent; processing the collected data by using the data fusion and analysis agent and the decision support agent, and transmitting the processed data to the dynamic strategy generation agent; the dynamic strategy generation agent outputs a corresponding service instruction according to the processed data, and interacts with the passenger cabin equipment control system in real time; according to the method, the passenger cabin can be managed in real time in an omnibearing manner, the existing limitation is broken, and aviation passenger cabin management is promoted to develop in an intelligent direction.

Description

Technical Field

[0001] The present invention relates to the field of big data processing, and particularly to a method and device for managing an aircraft cabin. Background Art

[0002] In modern air transportation, with the continuous improvement of passengers' requirements for flight experience and airlines' pursuit of operational efficiency, the traditional cabin management mode has been difficult to meet the actual needs. Currently, various types of data in the cabin are scattered and lack effective integration. Passenger behavior data has not been fully utilized to optimize services and operation models. At the same time, there is a lag in dealing with emergencies and equipment health management. Therefore, there is an urgent need for a platform that can comprehensively, real-time, and intelligently manage cabin human-machine-environment data to break the existing limitations and promote the development of aircraft cabin management towards the direction of intelligence. Summary of the Invention

[0003] To solve the above problems, the present invention provides a solution that can realize the optimization and upgrade of the cabin operation mode and provide accurate monitoring and efficient decision-making support for the on-board situation.

[0004] To achieve the above object, on the one hand, the present invention provides a method for managing an aircraft cabin, including: collecting data of the aircraft cabin through a passenger behavior perception intelligent agent, an environment monitoring intelligent agent, a device status monitoring intelligent agent, and a flight attendant physiological state monitoring intelligent agent respectively; using a data fusion and analysis intelligent agent and a decision support intelligent agent to process the collected data and then transmit it to a dynamic policy generation intelligent agent; the dynamic policy generation intelligent agent outputs corresponding service instructions according to the processed data and interacts with the cabin equipment control system in real time. Optionally, the passenger behavior perception intelligent agent includes: a micro-motion perception module based on UWB radar, which extracts passenger posture change data through Doppler characteristics; a visual attention mechanism analysis module, which uses a lightweight convolutional neural network to identify the fixation focus of passengers on the entertainment screen; a touch behavior modeling unit, which records the operation trajectory and generates a heat map Where λ is the attenuation coefficient, δ is the Dirac function, x is the abscissa of the passenger's real-time spatial trajectory, y is the ordinate of the passenger's real-time spatial trajectory, t is the time, t0 is the initial time, x t is the abscissa of the passenger's real-time spatial trajectory at time t, y t is the ordinate of the passenger's real-time spatial trajectory at time t.

[0005] Optionally, the passenger behavior perception intelligent agent adopts a non-contact sensor fusion technology, and generates a passenger preference vector through a multi-modal feature fusion algorithm F = α·V + β·A + γ·T, where V is the visual feature, A is the auditory feature, T is the tactile feature, and α + β + γ = 1.

[0006] Optionally, the environmental monitoring agent combines a micro environmental sensor array with digital twin technology, including: a thermo-humidity coupling sensor based on MEMS, with a sampling frequency ≥ 100 Hz and a measurement error < ±0.5%; a distributed fiber optic acoustic sensing system arranged along the cabin ceiling, with a spatial resolution ≤ 1 m; a dynamic comfort evaluation model: C = w1·f(T) + w2·g(H) + w3·h(L,N), where: f(T) = 1 / (1 + e^(-k(T - T0))) is the temperature sigmoid function, g(H) = |H - H0| / ΔH is the humidity deviation function, and h(L,N) = a·L^2 + b·N + c is the quadratic fitting function of light - noise.

[0007] Optionally, the environmental monitoring agent dynamically adjusts the weight coefficients [w1, w2, w3] through the particle swarm optimization algorithm, and updates the parameters every 5 minutes.

[0008] Optionally, the equipment status monitoring agent includes: a fault prediction network based on LSTM, with the input sequence X t = [Uf(t), Pf(t), At(t)], and the output fault probability P_f = σ(W h ·h t + b), where Uf(t) is, Pf(t) is, At(t) is, W h is, b is, h t is the hidden state, and σ is the sigmoid function; an adaptive threshold warning module that dynamically calculates the fault threshold θ = μ + 3σ based on the equipment historical data, where μ is the mean time between failures and σ is the standard deviation; a maintenance decision matrix D ij = argmax[ω1·R ij - ω2·C ij ], where R ij is the repair rate of the repair plan i for the fault j, C ij is the repair cost, and ω1 + ω2 = 1.

[0009] Optionally, the stewardess physiological state monitoring agent is integrated into the smart work ID device and realizes non-invasive monitoring by using flexible electronic skin technology, including: a heart rate variability analysis module based on PPG signals to calculate time domain indexes such as SDNN and RMSSD; a surface electromyogram signal acquisition unit to extract the energy characteristics of the 0 - 500 Hz frequency band through wavelet transform; a multi-task learning model: simultaneously outputting the fatigue level F ∈ [0, 4] and the work pressure index S ∈ [0, 100], sharing the feature extraction layer but having independent fully connected branches.

[0010] Optionally, the dynamic policy generation agent constructs a decision-making model based on double-delayed DDPG deep reinforcement learning, including: state space S = [passenger characteristics, environmental parameters, device status, crew status] ∈ R^128; action space A = {adjust temperature, recommend content, send reminder}, a total of 20 types of operations; reward function R = λ1·QoS + λ2·Efficiency - λ3·Energy, where: QoS = Σ(user satisfaction), Efficiency = 1 / (1 + response delay), Energy = normalized value of device power consumption; network update period τ = 0.01.

[0011] On the other hand, the present invention also provides a management device for an aircraft cabin, including: a data acquisition unit for respectively collecting data of the aircraft cabin through a passenger behavior perception agent, an environmental monitoring agent, a device status monitoring agent, and a crew physiological status monitoring agent, a data processing unit for using a data fusion and analysis agent and a decision support agent to process the collected data and then transmit it to the dynamic policy generation agent; an execution unit for outputting corresponding service instructions according to the processed data and interacting with the cabin equipment control system in real time.

[0012] The beneficial effects of the present invention compared with the prior art are: collecting cabin data through agents such as a passenger behavior perception agent, an environmental monitoring agent, a device status monitoring agent, and a crew physiological status monitoring agent, processing the data using a data fusion and analysis agent (DFAA) and a decision support agent, and finally outputting personalized service instructions through a dynamic policy generation agent (DSGA) and interacting with the cabin equipment control system in real time, which can manage the cabin in all aspects and in real time, break the existing limitations, and promote the development of aircraft cabin management towards the direction of intelligence. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 is a flowchart of a management method for an aircraft cabin provided by the present invention;

[0014] Figure 2 is a structural diagram of a management device for an aircraft cabin provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0015] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

[0016] Refer to Figure 1, this embodiment provides a method for managing an aircraft cabin, including the following steps:

[0017] S10: Collect data of the aircraft cabin through a passenger behavior perception agent, an environment monitoring agent, a device status monitoring agent, and a flight attendant physiological state monitoring agent respectively.

[0018] Specifically, connect various Internet of Things sensors through an on-board edge computing server, such as entertainment system usage sensors installed on the seat armrests, catering facility sensors located in the overhead bins, temperature and humidity sensors, noise sensors distributed in various areas of the cabin, and wearable devices worn by flight attendants, etc., to collect data on the usage of entertainment systems and catering facilities by passengers, changes in cabin environment parameters, equipment operation status, and flight attendant physiological state in real time.

[0019] In this embodiment, the Passenger Behavior Sensing Agent (PBSA) adopts a non-contact sensor fusion technology, and generates a passenger preference vector through a multi-modal feature fusion algorithm F = α·V + β·A + γ·T (V is visual feature, A is auditory feature, T is tactile feature, α + β + γ = 1).

[0020] Specifically, it includes: a micro-motion sensing module based on UWB radar, which extracts passenger posture change data through Doppler features; a visual attention mechanism analysis module, which uses a lightweight convolutional neural network (CNN) to identify the gaze focus of passengers on the entertainment screen; and a touch behavior modeling unit, which records the operation trajectory and generates a heat map H(x,y) = Σt = 1n e^(-λ(t - t0)) * δ(x - x t ,y - y t ), where λ is the attenuation coefficient and δ is the Dirac function.

[0021] In this embodiment, the Environment Monitoring Agent (EMA) combines a micro-environment sensor array with digital twin technology, and dynamically adjusts the weight coefficients [w1, w2, w3] through a particle swarm optimization algorithm, and updates the parameters every 5 minutes.

[0022] Specifically, it includes: a thermal-humidity coupling sensor based on MEMS, with a sampling frequency ≥ 100Hz and a measurement error < ±0.5%; and a distributed fiber optic acoustic sensing (DAS) system, arranged along the cabin ceiling, with a spatial resolution ≤ 1m; and a dynamic comfort evaluation model: C = w1·f(T) + w2·g(H) + w3·h(L,N), where: f(T) = 1 / (1 + e^(-k(T - T0))) is the temperature sigmoid function, g(H) = |H - H0| / ΔH is the humidity deviation function, and h(L,N) = a·L^2 + b·N + c is the light-noise quadratic fitting function.

[0023] In this embodiment, the Equipment Status Monitoring Agent (ESMA) constructs a digital twin model of equipment health and realizes bidirectional data interaction with the Aircraft Maintenance System (AMS) through the OPC UA protocol.

[0024] Specifically, it includes: a fault prediction network based on LSTM, with the input sequence X t = [Uf(t), Pf(t), At(t)], and the output fault probability P_f = σ(W h · h t + b), where h t is the hidden state and σ is the sigmoid function; an adaptive threshold warning module that dynamically calculates the fault threshold θ = μ + 3σ based on the equipment historical data, where μ is the mean time between failures and σ is the standard deviation; a maintenance decision matrix D ij = argmax[ω1 · R ij - ω2 · C ij ], where R ij is the repair rate of the repair plan i for the fault j, C ij is the repair cost, and ω1 + ω2 = 1.

[0025] In this embodiment, the Cabin Crew Physiological State Monitoring Agent (CPSMA) is integrated into the smart work badge device and realizes non-intrusive monitoring by using flexible electronic skin technology.

[0026] Specifically, it includes: a heart rate variability (HRV) analysis module based on PPG signals to calculate time domain indexes such as SDNN and RMSSD; a surface electromyography (sEMG) signal acquisition unit to extract energy features in the 0 - 500Hz frequency band through wavelet transform; a multi-task learning model: simultaneously outputting the fatigue level F ∈ [0, 4] and the work pressure index S ∈ [0, 100], sharing a feature extraction layer but having independent fully connected branches.

[0027] S20: Use the data fusion and analysis agent and the decision support agent to process the collected data and then transfer it to the dynamic policy generation agent.

[0028] Specifically, first preprocess the collected data, perform operations such as removing noise, normalizing the data, and filling in missing values to ensure the accuracy and integrity of the data. Then fuse the processed data, integrating data from different sources and of different types into a unified data model. Next, use deep learning algorithms to analyze the fused data, mine the behavior patterns and preferences of passengers, and predict the cabin operation trends in the next period of time, such as the peak of catering demand, the popularity of entertainment content, etc.

[0029] In this embodiment, the data fusion and analysis agent adopts a federated learning framework of spatiotemporal joint coding, including: spatial encoder: using graph convolutional network (GCN) to model the topological relationship between agents; temporal encoder: using Transformer architecture to extract long-term dependency features of data sequences; model aggregation mechanism: updating the global model parameter θ through dynamic weighted average algorithm global =Σ(α i ·θ local_i ), where the weight α i =softmax(β·ACC i ), β is the temperature coefficient; the framework supports differential privacy protection and adds Gaussian noise ε~N(0,σ^2I) when updating the gradient.

[0030] S30: The dynamic strategy generation agent outputs corresponding service instructions according to the processed data, and interacts with the cabin equipment control system in real time.

[0031] Specifically, the dynamic strategy generation agent generates personalized service strategies, fleet operation suggestions, and flight allocation plans based on data. For example, based on the historical behavior data and real-time feedback of passengers, a customized service plan is provided for each passenger, such as recommending entertainment content that suits their preferences, preparing their preferred meals in advance, etc. At the same time, the operation data of each flight is analyzed, including indicators such as passenger satisfaction and cost-effectiveness, to provide scientific and reasonable suggestions for the route planning and scheduling of the fleet. In addition, based on the actual situation in the cabin and external weather, airspace and other information, the parameters such as flight take-off and landing time and flight altitude are optimized to improve the overall operational efficiency of the flight. In this embodiment, the dynamic strategy generation agent (DSGA) constructs a decision model based on double-delay DDPG deep reinforcement learning, including: state space S = [passenger characteristics, environmental parameters, equipment status, crew status] ∈ R^128; action space A = {adjust temperature, recommend content, send reminders...} a total of 20 types of operations; reward function R = λ1·QoS+λ2·Efficiency-λ3·Energy, where: QoS = Σ(user satisfaction), Efficiency = 1 / (1+response delay), Energy = standardized value of device power consumption; network update period τ = 0.01.

[0032] In another embodiment, when an abnormal situation is monitored, such as excessive smoke concentration, precursors to equipment failure, excessive fatigue of crew members, etc., an alarm signal is immediately triggered, and the relevant information is promptly notified to the crew and the ground control center so that appropriate measures can be taken to deal with it.

[0033] In this embodiment, a distributed trust mechanism based on blockchain technology is used between intelligent entities to verify data, thereby improving the security and privacy of data.

[0034] In some other embodiments, various types of data and analysis results in the cabin can be presented in the form of intuitive charts, maps, etc., such as the passenger distribution heat map, equipment status dashboard, etc., to facilitate users to quickly obtain key information. At the same time, it supports users to filter and query data according to different dimensions such as time, location, passenger type, etc., to deeply understand the cabin operation under specific conditions. In addition, an intelligent decision-making assistance function is also provided, which displays the simulation effects of different operation strategies through a visual interface to help decision-makers make more scientific and reasonable decisions. This management solution collects cabin data through agents such as the passenger behavior perception agent, environment monitoring agent, equipment status monitoring agent, and flight attendant physiological state monitoring agent, processes the data using the data fusion and analysis agent (DFAA) and decision support agent, and finally outputs personalized service instructions through the dynamic strategy generation agent (DSGA) and interacts with the cabin equipment control system in real time, enabling all-round real-time management of the cabin, breaking the existing limitations, and promoting the development of airline cabin management towards the direction of intelligence. Refer to Figure 2 In addition, this embodiment also provides a management device for an aircraft cabin, including:

[0035] A data acquisition unit 100, configured to collect data of the aircraft cabin through a passenger behavior perception agent, an environment monitoring agent, an equipment status monitoring agent, and a flight attendant physiological state monitoring agent respectively; it should be noted that since the specific data acquisition methods and processes have been elaborated in detail in step S10 of the above-mentioned management method for an aircraft cabin, they will not be repeated here.

[0036] A data processing unit 200, configured to process the collected data using a data fusion and analysis agent and a decision support agent and then transfer it to a dynamic strategy generation agent; it should be noted that since the specific data processing methods and processes have been elaborated in detail in step S20 of the above-mentioned management method for an aircraft cabin, they will not be repeated here.

[0037] An execution unit 300, configured to output corresponding service instructions according to the processed data and interact with the cabin equipment control system in real time. It should be noted that since the specific execution methods and processes have been elaborated in detail in step S30 of the above-mentioned management method for an aircraft cabin, they will not be repeated here.

[0038] In addition, an embodiment of the present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium can store a program, and when the program is executed, it includes some or all of the steps of any management method for an aircraft cabin recorded in the above method embodiments.

[0039] In addition, in each embodiment of the present invention, each functional unit can be integrated into a processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of a software functional unit. If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable memory. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a memory and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned memory includes: USB flash drives, read-only memories (ROM, Read-Only Memory), random access memories (RAM, RandomAccess Memory), mobile hard disks, magnetic disks, or optical disks and other media that can store program codes.

[0040] Those of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing relevant hardware through a program. The program can be stored in a computer-readable memory. The memory can include: flash drives, read-only memories (English: Read-Only Memory, abbreviated as: ROM), random access memories (English: RandomAccess Memory, abbreviated as: RAM), magnetic disks, or optical disks, etc. The above has described an exemplary flowchart for implementing an aircraft cabin management method according to an embodiment of the present invention with reference to the accompanying drawings. It should be noted that the large number of details included in the above description are only an exemplary illustration of the present invention, rather than a limitation of the present invention. In other embodiments of the present invention, the method may have more, fewer, or different steps, and the order, inclusion, function, and other relationships between the steps may be different from those described and illustrated.

Claims

1. A management method for an aircraft cabin, characterized in that Including: Collecting data of the aircraft cabin through a passenger behavior perception agent, an environment monitoring agent, a device status monitoring agent, and a flight attendant physiological state monitoring agent respectively; Using a data fusion and analysis agent and a decision support agent to process the collected data and then transmit it to a dynamic policy generation agent; The dynamic policy generation agent outputs corresponding service instructions according to the processed data and interacts with the cabin equipment control system in real time.

2. The management method according to claim 1, characterized in that, The passenger behavior perception agent includes: A micro-motion perception module based on UWB radar, which extracts passenger posture change data through Doppler features; a visual attention mechanism analysis module, which uses a lightweight convolutional neural network to identify the gaze focus of passengers on the entertainment screen; Touch behavior modeling unit, records the operation trajectory and generates a heat map where λ is the attenuation coefficient, δ is the Dirac function, x is the abscissa of the real-time spatial trajectory of the passenger, y is the ordinate of the real-time spatial trajectory of the passenger, t is the time, t0 is the initial time, x t is the abscissa of the real-time spatial trajectory of the passenger at time t, y t is the ordinate of the real-time spatial trajectory of the passenger at time t.

3. The management method according to claim 2, wherein: The passenger behavior perception agent adopts a non-contact sensor fusion technology and generates a passenger preference vector through a multi-modal feature fusion algorithm F = α·V + β·A + γ·T, where V is the visual feature, A is the auditory feature, T is the tactile feature, and α + β + γ = 1.

4. The management method according to claim 1, characterized in that, The environment monitoring agent combines a micro environmental sensor array with digital twin technology, including: A thermal-humidity coupling sensor based on MEMS, with a sampling frequency ≥ 100Hz and a measurement error < ±0.5%; A distributed fiber optic acoustic sensing system arranged along the cabin ceiling, with a spatial resolution ≤ 1m; A dynamic comfort evaluation model: C = w1·f(T) + w2·g(H) + w3·h(L,N), where: f(T) = 1 / (1 + e^(-k(T - T0))) is the temperature sigmoid function, g(H) = |H - H0| / ΔH is the humidity deviation function, and h(L,N) = a·L^2 + b·N + c is the light-noise quadratic fitting function.

5. The management method according to claim 4, wherein: The environment monitoring agent dynamically adjusts the weight coefficients [w1, w2, w3] through a particle swarm optimization algorithm and updates the parameters every 5 minutes.

6. The management method according to claim 1, wherein The device status monitoring agent includes: LSTM-based fault prediction network, with input sequence X t = [Uf(t), Pf(t), At(t)], and the output fault probability P_f = σ(W h ·h t + b), where Uf(t) is the usage frequency of the device, Pf(t) is the historical fault probability, At(t) is the cumulative operation time, W h is the weight matrix, b is the bias correction term, h t is the hidden state, and σ is the sigmoid function; An adaptive threshold warning module, which dynamically calculates the fault threshold θ = μ + 3σ according to the device historical data, where μ is the mean time between failures and σ is the standard deviation; Maintain decision matrix D ij = argmax[ω1·R ij - ω2·C ij , where R ij is the repair rate of maintenance plan i for fault j, C ij is the maintenance cost, and ω1 + ω2 = 1.

7. The management method according to claim 1, characterized in that The flight attendant physiological state monitoring agent is integrated into the intelligent work ID device and realizes non-intrusive monitoring by using flexible electronic skin technology, including: A heart rate variability analysis module based on PPG signals, which calculates time domain indexes such as SDNN and RMSSD; A surface electromyogram signal acquisition unit, which extracts the energy characteristics of the 0 - 500Hz frequency band through wavelet transform; A multi-task learning model: simultaneously outputs the fatigue level F ∈ [0, 4] and the work pressure index S ∈ [0, 100], sharing a feature extraction layer but having independent fully connected branches.

8. The management method according to claim 1, wherein The dynamic policy generation agent constructs a decision model based on double-delay DDPG deep reinforcement learning, including: The state space S = [passenger characteristics, environmental parameters, device status, flight attendant status] ∈ R^128; The action space A = {adjust temperature, recommend content, send reminder}, a total of 20 types of operations; The reward function \(R = \lambda_1\cdot QoS+\lambda_2\cdot Efficiency-\lambda_3\cdot Energy\), where: \(QoS=\sum\) (user satisfaction), \(Efficiency = 1 / (1 + \text{response latency})\), and \(Energy\) is the normalized value of device power consumption; the network update period \(\tau = 0.01\).

9. A management device for an aircraft cabin, characterized in that, It includes: A data acquisition unit for collecting data of the aircraft cabin through a passenger behavior perception agent, an environment monitoring agent, a device status monitoring agent, and a flight attendant physiological state monitoring agent respectively, and a data processing unit for processing the collected data using a data fusion and analysis agent and a decision support agent and then transmitting it to a dynamic policy generation agent; An execution unit for outputting corresponding service instructions according to the processed data and interacting with the cabin equipment control system in real time.

10. A computer-readable storage medium, characterized in that, It includes: The computer-readable storage medium stores a computer program, characterized in that when the computer program is executed by a processor, it implements the steps of a method for managing an aircraft cabin according to any one of claims 1 to 8.

Images