Search

CN-121978937-A - Intelligent hotel multi-equipment integrated linkage control method and system

CN121978937ACN 121978937 ACN121978937 ACN 121978937ACN-121978937-A

Abstract

The invention relates to the technical field of intelligent control and discloses a multi-equipment integrated linkage control method and system for an intelligent hotel, wherein the method comprises the steps of collecting operation data and environment parameters of intelligent hotel equipment, and carrying out feature extraction and semantic fusion to obtain an equipment state feature vector sequence; the method comprises the steps of constructing a device relation map by adopting a data-driven learning method, identifying user scene labels and confidence by adopting a hierarchical classification model, outputting an optimized linkage control scheme by adopting a hierarchical optimization framework and a multi-objective optimization method, outputting an abnormal detection result and an emergency linkage control instruction by adopting a multi-parameter abnormal detection method and a hierarchical emergency response mechanism, obtaining an updated strategy model by adopting simulated learning pre-training and reinforcement learning on-line fine adjustment, executing control by adopting a distributed framework and a priority scheduling mechanism coordinated with edge cloud, and outputting an execution result. The intelligent linkage control method and the intelligent linkage control system can realize intelligent linkage control of multiple devices in the hotel and improve user experience and energy efficiency.

Inventors

  • WANG HUAIRUI
  • ZHOU YUGUO
  • WANG SHANPENG
  • Xia le
  • SHI WEIWEI

Assignees

  • 江苏信艺装饰工程有限公司

Dates

Publication Date
20260505
Application Date
20260130

Claims (10)

  1. 1. The intelligent multi-equipment integrated linkage control method for the hotel is characterized by comprising the following steps of: S1, acquiring operation data and environmental parameters of hotel intelligent equipment, and performing feature extraction and semantic fusion to obtain an equipment state feature vector sequence; s2, acquiring a device state feature vector sequence, and acquiring a device relation graph by adopting a data driving learning method; s3, according to the equipment state feature vector sequence, a hierarchical classification model is adopted to obtain a user scene label and a confidence coefficient; S4, receiving the equipment relation map, the user scene labels and the confidence coefficient, and outputting an optimized linkage control scheme by adopting a layered optimization framework and a multi-objective optimization method; S5, according to the equipment state feature vector sequence and the linkage control scheme, a multi-parameter abnormality detection method and a hierarchical emergency response mechanism are adopted, and an abnormality detection result report, an abnormality level classification, an emergency linkage control instruction and an influence range analysis are output; s6, acquiring feedback data after the linkage control scheme is executed, and adopting simulated learning pre-training and reinforcement learning to perform online fine adjustment to obtain an updated strategy model; and S7, receiving a linkage control scheme, an emergency linkage control instruction and an updated strategy model, and obtaining a control execution result, a system running state report and a data synchronization log by adopting a distributed architecture and a priority scheduling mechanism of edge cloud cooperation.
  2. 2. The intelligent hotel multi-equipment integrated linkage control method according to claim 1, wherein the S1 comprises: Configuring a multi-protocol-adaptive device data acquisition interface, developing an adapter aiming at communication protocols of different types of intelligent devices, establishing a unique identifier, a device type and a communication protocol type of a device registry recording device, and enabling a data acquisition agent to periodically acquire state data for each device creation data acquisition thread according to the registry configuration to obtain an original device state data stream; And executing time synchronization alignment of the multi-source data, selecting a uniform time reference and a sampling interval to establish a discrete time point sequence, adopting an aggregation method in a time window for the high-frequency acquired equipment data, and adopting an interpolation method for the low-frequency acquired equipment data to fill in so as to obtain an equipment state snapshot sequence on the uniform time reference.
  3. 3. The intelligent hotel multi-equipment integrated linkage control method according to claim 1, wherein the step S2 comprises: Abstracting each intelligent device in the hotel into one node in the graph, establishing an initial association edge between the device nodes based on the device topology configuration file, dividing the association edge into a physical adjacent edge, a function dependent edge and a manufacturer recommended edge, and giving initial weights to the edges of different types to construct an initial device relation graph; Performing representation learning on the equipment relation graph by adopting a graph neural network model, aggregating neighbor information of nodes through multi-layer graph convolution operation, and introducing an attention mechanism into the graph neural network to automatically learn importance weights of the neighbor nodes; training the graphic neural network model by using the historical operation data, designing self-supervision learning of equipment state prediction tasks and equipment association prediction tasks, generating equipment nodes embedded representation of the trained graphic model, and learning linkage modes among equipment.
  4. 4. The intelligent hotel multi-equipment integrated linkage control method according to claim 1, wherein the step S3 comprises: constructing a multi-mode feature extraction module to respectively encode equipment interaction sequence data, user position track data, time context information and environmental parameter data; Designing a cross-mode attention fusion mechanism, mapping feature vectors of all modes to the same dimensional space, calculating correlation scores among features of different modes, obtaining an attention weight matrix through normalization processing, and carrying out weighted summation on value vectors of all modes by using the attention weight to obtain a fused feature representation; And establishing a hierarchical scene recognition model, judging whether a user is in a guest room and in an activity state by a first-stage classifier, judging a specific activity type by a second-stage classifier on the basis of a first-stage classification result, and outputting scene categories and confidence degrees by the model.
  5. 5. The intelligent hotel multi-equipment integrated linkage control method according to claim 1, wherein the step S4 comprises: Inquiring a scene knowledge base according to a user scene tag to acquire a target environment parameter range and equipment control preference under the scene, retrieving historical preference data of a user and fusing personalized preference into target parameters; Decomposing the decision-making problem into a strategy layer and a tactical layer by adopting a hierarchical optimization framework, wherein the strategy layer determines a device set and a target environment state which participate in linkage regulation and control, and the tactical layer generates specific control parameters and execution time sequences for each device in a device subset; The strategic layer executes a device screening algorithm based on the device relation graph, performs extended search along the associated edges from the core device nodes, and judges whether neighbor devices participate in linkage according to the types and weights of the edges to obtain a device subset strongly related to the current scene; The strategic layer establishes a multi-objective optimization model, and the objective function comprises maximizing user comfort, minimizing energy consumption and minimizing response time, and solving the optimization model by adopting a multi-objective evolutionary algorithm to obtain a target environment state vector.
  6. 6. The intelligent hotel multi-equipment integrated linkage control method according to claim 1, wherein the step S5 comprises: establishing a state prediction model for key equipment to perform short-term prediction on the operation parameters of the equipment, wherein the input of the state prediction model comprises the historical state time sequence of the equipment, the state of related equipment, a current control instruction and environmental conditions; Designing a self-adaptive threshold mechanism to dynamically adjust an abnormality detection threshold according to the current working state and the historical fluctuation characteristic of the equipment; predicting the operation parameters of the equipment by using a state prediction model, acquiring residual errors between the actual observation parameters of the equipment and the predicted values, carrying out standardization processing on the residual errors, and judging that the equipment enters an abnormal state when the comprehensive abnormal score of the equipment exceeds a comprehensive judgment threshold value; And establishing an abnormality severity assessment model to comprehensively consider the amplitude, duration, influence range and potential risk of the abnormality to divide the abnormal event into low-level abnormality, medium-level abnormality and high-level abnormality, and triggering an emergency response mechanism of a corresponding level according to the abnormality level assessment result.
  7. 7. The intelligent hotel multi-equipment integrated linkage control method according to claim 1, wherein the step S6 comprises: establishing a multi-dimensional evaluation system quantitative control quality, wherein the evaluation system comprises objective indexes and subjective indexes, carrying out normalization treatment on the objective indexes and the subjective indexes, and obtaining a comprehensive evaluation score of the linkage effect through weighted combination; Training a strategy network by adopting a supervised learning method to minimize the difference between the action output by the network and the expert action; establishing a reinforcement learning framework to model a coordinated control strategy optimization problem as a Markov decision process, and defining a state space, an action space and a reward function; The strategy network is trained by adopting a strategy gradient method, the strategy network and the value network are trained simultaneously by adopting an actor-critter architecture in the training process, and the gap between the new strategy and the old strategy is limited when the strategy is updated by adopting a near-end strategy optimization algorithm.
  8. 8. The intelligent hotel multi-equipment integrated linkage control method according to claim 2, wherein the step S1 further comprises: Designing a data caching and compensating mechanism to solve the problems of equipment offline and data missing, maintaining a local data caching queue for each equipment in a data acquisition agent, judging that the equipment enters an offline state to trigger a state estimation compensating process when the equipment is detected to be continuously acquired for multiple times or response is overtime, and assigning confidence scores to the compensated estimated data by adopting a nearest neighbor filling method, a time sequence extrapolation method or a state reasoning method based on a correlation model according to different conditions; Extracting multi-time scale equipment state features to construct vectorized representation of equipment states, wherein the feature extraction considers instantaneous features, short-term statistical features, long-term mode features and cross features among equipment, and various extracted features are combined to form comprehensive feature vectors of the equipment states and standardized.
  9. 9. The intelligent hotel multi-equipment integrated linkage control method according to claim 1, wherein the step S7 comprises: the method comprises the steps that a distributed control architecture with cooperative edge computing and cloud computing is adopted, an edge computing server is deployed locally in a hotel to take charge of tasks with high real-time requirements, a cloud deployment management platform is used for taking charge of computation-intensive tasks, an edge node periodically uploads equipment state data and control logs to the cloud, and the cloud transmits updated model parameters and optimized strategies to the edge node; Establishing a priority scheduling mechanism to perform conflict detection and priority sequencing on the multi-source control instructions, defining the priority level of the control instructions, and executing conflict detection when a plurality of instructions aim at the same device to execute high-priority instructions according to priority selection; The transaction mechanism for designing equipment control encapsulates a group of control instructions needing to be cooperatively executed into a transaction to ensure the atomicity and consistency of linkage control.
  10. 10. An intelligent multi-equipment integrated linkage control system for a hotel, which is used for executing the steps in the intelligent multi-equipment integrated linkage control method for the hotel according to any one of claims 1 to 9, and is characterized by comprising the following steps: the data acquisition and feature extraction module is used for acquiring operation data and environment parameters of the hotel intelligent equipment and carrying out feature extraction and semantic fusion to obtain an equipment state feature vector sequence; The device relation graph construction module is used for acquiring a device state feature vector sequence and acquiring a device relation graph by adopting a data driving learning method; the scene recognition module is used for obtaining a user scene tag and a confidence coefficient by adopting a hierarchical classification model according to the equipment state feature vector sequence; The linkage decision optimization module is used for receiving the equipment relation graph, the user scene label and the confidence coefficient, adopting a layered optimization framework and combining a multi-objective optimization method, and outputting an optimization linkage control scheme; the abnormality detection and emergency response module is used for outputting an abnormality detection result report, abnormality grade classification, emergency linkage control instructions and influence range analysis by adopting a multi-parameter abnormality detection method and a grading emergency response mechanism according to the equipment state feature vector sequence and the linkage control scheme; the strategy learning and optimizing module is used for acquiring feedback data after the linkage control scheme is executed, and adopting simulated learning pre-training and reinforcement learning on-line fine adjustment to obtain an updated strategy model; The distributed execution and monitoring module is used for receiving the linkage control scheme, the emergency linkage control instruction and the updated strategy model, and obtaining control execution results, a system running state report and a data synchronization log by adopting an edge cloud cooperative distributed architecture and a priority scheduling mechanism.

Description

Intelligent hotel multi-equipment integrated linkage control method and system Technical Field The invention relates to the technical field of intelligent control, in particular to a multi-equipment integrated linkage control method and system for an intelligent hotel. Background With the rapid development of the internet of things technology and the artificial intelligence technology, intelligent hotels have become an important direction of transformation and upgrading in the hotel industry. A large number of intelligent devices including intelligent door locks, lighting systems, air conditioning systems, curtain systems, environmental sensors, etc. are deployed in modern hotel rooms, which can provide more convenient and comfortable check-in experience for guests. However, most of the existing hotel intelligent systems adopt single-device independent control or simple scene mode switching, and lack deep coordination and intelligent linkage capability between devices. The method mainly comprises the steps that data acquisition and fusion processing capacity of multi-source heterogeneous equipment is insufficient, equipment of different manufacturers adopts different communication protocols and data formats, unified data acquisition and state monitoring are difficult to achieve, information island among the equipment is caused, secondly, identification accuracy of a user behavior scene is low, an existing system mainly depends on simple rule judgment or a single data source, real requirements and behavior intention of a user cannot be accurately understood, misjudgment easily occurs, control strategies are improper, optimizing capacity of equipment linkage control strategies is limited, the existing method mostly adopts a preset fixed scene mode, global optimization of multi-equipment cooperative control is lacked, and optimal balance among user comfort, energy efficiency and response speed cannot be achieved. In addition, the existing system has insufficient capability in the aspects of abnormality detection and emergency response, and when equipment fails or is abnormal, the system is difficult to discover and take effective compensation measures in time, so that the service quality and the user experience are affected. Meanwhile, the existing system lacks continuous learning and self-adaptive optimization mechanisms, can not continuously improve control strategies according to user feedback and operation data, and is difficult to adapt to individual demands of different users. Disclosure of Invention The invention provides an intelligent multi-equipment integrated linkage control method and system for hotels, which solve the technical problems of low accuracy of user scene identification, limited optimization capability of equipment linkage control strategies, abnormal detection and insufficient emergency response capability in the related technology. The invention provides a multi-equipment integrated linkage control method for an intelligent hotel, which comprises the following steps: S1, acquiring operation data and environmental parameters of hotel intelligent equipment, and performing feature extraction and semantic fusion to obtain an equipment state feature vector sequence; s2, acquiring a device state feature vector sequence, and acquiring a device relation graph by adopting a data driving learning method; s3, according to the equipment state feature vector sequence, a hierarchical classification model is adopted to obtain a user scene label and a confidence coefficient; S4, receiving the equipment relation map, the user scene labels and the confidence coefficient, and outputting an optimized linkage control scheme by adopting a layered optimization framework and a multi-objective optimization method; S5, according to the equipment state feature vector sequence and the linkage control scheme, a multi-parameter abnormality detection method and a hierarchical emergency response mechanism are adopted, and an abnormality detection result report, an abnormality level classification, an emergency linkage control instruction and an influence range analysis are output; s6, acquiring feedback data after the linkage control scheme is executed, and adopting simulated learning pre-training and reinforcement learning to perform online fine adjustment to obtain an updated strategy model; and S7, receiving a linkage control scheme, an emergency linkage control instruction and an updated strategy model, and obtaining a control execution result, a system running state report and a data synchronization log by adopting a distributed architecture and a priority scheduling mechanism of edge cloud cooperation. In a preferred embodiment, the S1 includes: Configuring a multi-protocol-adaptive device data acquisition interface, developing an adapter aiming at communication protocols of different types of intelligent devices, establishing a unique identifier, a device type and a communication protocol type of a device reg