CN-121051653-B - Intelligent space operation and maintenance optimization method and device based on data elements

CN121051653BCN 121051653 BCN121051653 BCN 121051653BCN-121051653-B

Abstract

The invention relates to the technical field of intelligent space operation and maintenance and discloses a method and a device for intelligent space operation and maintenance based on data elements, wherein the method comprises the steps of collecting operation data of an Internet of things terminal, an edge node and a cloud platform in an intelligent space, and fusing the operation data to obtain fused data; the intelligent space abnormal situation report is obtained by carrying out sensitive information identification and encryption on the fusion data, obtaining encrypted data, inputting the encrypted data into local models of all edge nodes, extracting equipment operation characteristics, integrating equipment operation characteristic sets in a cloud aggregation server, carrying out abnormality identification based on the equipment operation characteristic sets to obtain equipment abnormal state information, carrying out abnormality cause tracing analysis according to the equipment abnormal state information, and providing an interpretable abnormal analysis path and a comprehensive situation perception report for operation and maintenance personnel.

Inventors

XUE YUERONG
XIE JINLONG
OUYANG SHIXIONG

Assignees

深圳市金地楼宇科技有限公司

Dates

Publication Date: 20260505
Application Date: 20250926

Claims (7)

1. The intelligent space operation and maintenance optimization method based on the data elements is characterized by comprising the following steps of: Acquiring operation data of an Internet of things terminal, an edge node and a cloud platform in an intelligent space, and fusing the operation data to obtain fused data, wherein the operation data comprises the steps of acquiring respective operation data from the Internet of things terminal, the edge node and the cloud platform in the intelligent space, calculating terminal edge similarity of the Internet of things terminal data and the edge node data in the operation data, calculating edge cloud similarity and equipment time coordination of the edge node data and the cloud platform data in the operation data, and carrying out association fusion based on the terminal edge similarity, the edge cloud similarity and the equipment time coordination to obtain fused data; sensitive information identification and encryption are carried out on the fusion data, so that encrypted data are obtained; Inputting the encrypted data into local models of all edge nodes, extracting equipment operation characteristics, and integrating equipment operation characteristic sets in a cloud aggregation server; The method comprises the steps of carrying out anomaly identification based on an equipment operation feature set to obtain equipment abnormal state information, specifically converting the equipment operation feature set into equipment state feature vectors comprising a temperature state, a power state, a network state and a time state, processing the equipment state feature vectors by using an isolated forest algorithm, constructing a binary tree isolation structure by random sampling, calculating the path depth of the equipment state feature vectors to obtain equipment operation anomaly score, analyzing a historical distribution rule of the equipment operation anomaly score by a training reinforcement learning model, dynamically optimizing a normal state boundary and an anomaly judgment critical value by combining expert experience labeling data to obtain a self-adaptive anomaly threshold value, and comparing the equipment operation anomaly score with the self-adaptive anomaly threshold value, identifying equipment exceeding the normal boundary and labeling anomaly type and severity level to obtain equipment anomaly state information; The method comprises the steps of carrying out anomaly cause tracing analysis according to equipment anomaly state information to obtain a smart space anomaly situation report, specifically, retrieving a knowledge graph of an anomaly event of the smart space equipment, matching equipment types and anomaly modes in the equipment anomaly state information, creating anomaly tracing basic data based on the equipment types and the anomaly modes, operating a thinking chain reasoning engine to analyze the anomaly tracing basic data step by step to obtain a reasoning tracing result, accessing an LSM (least squares) tree log storage system to inquire historical operation records of the smart space equipment, identifying historical anomaly modes which have relevance with current anomaly events in time and space to obtain anomaly event space-time correlation data, and integrating the reasoning tracing result and the anomaly event space-time correlation data to generate the smart space anomaly situation report.
2. The intelligent space operation and maintenance optimization method based on data elements according to claim 1, wherein the calculating the edge cloud similarity and the equipment time coordination of the edge node data and the cloud platform data in the operation data comprises: extracting the numerical range, the data type and the change trend characteristics of the edge node data from the operation data, constructing an edge node characteristic vector, extracting corresponding characteristic parameters of cloud platform data, and constructing a cloud platform characteristic vector; Cosine similarity calculation is carried out on the edge node feature vector and the cloud platform feature vector, and edge cloud similarity is obtained; Establishing a sliding time window based on the timestamp information of the operation data, and analyzing the arrival time difference and the data updating frequency of the edge node data and the cloud platform data in the same sliding time window to obtain a time synchronization analysis result; and calculating the time delay variance and the frequency matching degree of the edge node data and the cloud platform data transmission according to the time synchronization analysis result, and calculating the equipment time coordination based on the time delay variance and the frequency matching degree.
3. The intelligent space operation and maintenance optimization method based on data elements according to claim 1, wherein the performing sensitive information identification and encryption on the fusion data to obtain encrypted data comprises: Establishing a sensitive data tag system according to the fusion data, wherein the sensitive data tag system comprises a public level, an internal level, a secret level and an absolute secret level; according to the sensitive data tag system, the fusion data is input into a LoRA fine-tuned large language model for intelligent classification to obtain intelligent classification data, wherein the intelligent classification data comprises public level data, internal level data, secret level data and secret level data; And (3) using an SM4 symmetric encryption algorithm for the public-level data and the internal-level data, and using an SM2 digital signature and an SM3 integrity check sum SM4 for the secret-level data and the absolute-level data in combination to obtain encrypted data.
4. The method for optimizing intelligent space operation and maintenance based on data elements according to claim 1, wherein the steps of extracting the device operation feature from the local model of the encrypted data input to each edge node, and integrating the device operation feature set in the cloud aggregation server include: Distributing the encrypted data to a local model of each edge node, and extracting equipment operation characteristics from the temperature, humidity, current, voltage and power data of the encrypted state based on the local model; The homomorphic encryption algorithm is adopted to encrypt and protect the running characteristics of the equipment and the weight parameters of the local model, so that the characteristic data is ensured to always keep a ciphertext state in network transmission, and homomorphic encrypted characteristic data is obtained; Uploading the homomorphic encryption characteristic data of each edge node to a cloud aggregation server, and executing homomorphic addition operation and homomorphic multiplication operation to realize weighted aggregation calculation in a ciphertext state, so as to obtain global aggregation characteristics; updating the global model parameters of federal learning based on the global aggregation features to obtain updated model parameters, issuing the updated model parameters to each edge node, and combining the equipment operation features of each edge node to obtain an equipment operation feature set.
5. The method for optimizing operation and maintenance of intelligent space based on data elements according to claim 4, wherein updating the global model parameters of federal learning based on the global aggregation feature to obtain updated model parameters, and issuing the updated model parameters to each edge node and combining the device operation features of each edge node to obtain a device operation feature set, comprises: calculating a loss function gradient of the federal learning model based on the global aggregation feature, and updating a weight matrix and a bias vector of the local model by adopting a federal average algorithm based on the loss function gradient to obtain updated model parameters; transmitting the updated model parameters from the cloud to each edge node through an encryption communication protocol, and synchronously updating a weight matrix and a bias vector of a local model by each edge node to obtain an edge node model with synchronous parameters; Operating the edge node model with the synchronous parameters to re-extract the equipment operation characteristics in the encrypted data to obtain the synchronized equipment operation characteristics; And converging the synchronized equipment operation characteristics of each edge node to obtain an equipment operation characteristic set.
6. The data element-based smart space operation and maintenance optimization method of claim 1, further comprising: Extracting abnormal equipment position, fault type and severity information in the intelligent space abnormal situation report, constructing three-dimensional modeling of the intelligent space physical environment, marking an abnormal area and equipment nodes, and obtaining a three-dimensional abnormal situation map; Evaluating risk propagation paths and influence ranges of various abnormal areas in the three-dimensional abnormal situation map, and calculating area risk grade data; matching the regional risk level data with candidate operation and maintenance processing strategies in a historical operation and maintenance case library; And rendering the candidate operation and maintenance processing strategy to an interactive three-dimensional interface, and displaying the risk level, dynamically labeling the processing step and integrating the mobile terminal control function by using color coding to obtain an intelligent space comprehensive operation and maintenance decision scheme.
7. A data element-based smart space operation and maintenance optimization apparatus for performing the data element-based smart space operation and maintenance optimization method of any one of claims 1 to 6, the data element-based smart space operation and maintenance optimization apparatus comprising: The intelligent space acquisition module is used for acquiring operation data of the internet of things terminal, the edge node and the cloud platform in the intelligent space and fusing the operation data to obtain fused data; the encryption module is used for carrying out sensitive information identification and encryption on the fusion data to obtain encrypted data; The extraction module is used for inputting the encrypted data into the local model of each edge node to extract the equipment operation characteristics, and integrating the equipment operation characteristic sets in the cloud aggregation server; the abnormality identification module is used for carrying out abnormality identification based on the equipment operation characteristic set to obtain equipment abnormal state information; And the traceability analysis module is used for carrying out traceability analysis of the abnormal reasons according to the equipment abnormal state information to obtain an intelligent space abnormal situation report.

Description

Intelligent space operation and maintenance optimization method and device based on data elements Technical Field The invention relates to the technical field of intelligent space operation and maintenance, in particular to an intelligent space operation and maintenance optimization method and device based on data elements. Background With the rapid development of the internet of things, artificial intelligence and cloud computing technologies, smart space is an important component of smart city construction, and has been widely applied to scenes such as smart buildings, smart communities, smart parks and the like. The intelligent space realizes real-time acquisition and monitoring of multidimensional information such as environmental parameters, equipment states, energy consumption data and the like by deploying a large number of terminal equipment of the Internet of things, edge computing nodes and cloud platform services. However, the existing intelligent space operation and maintenance management technology generally has the problem of data island, and an effective data association and fusion mechanism is lacking between various devices and systems, so that unified operation and maintenance situation awareness capability is difficult to form. The traditional intelligent space operation and maintenance method mainly relies on a single data source to carry out simple threshold comparison and rule matching, and cannot effectively process complex association relations of multi-source heterogeneous data, so that the problems of low abnormality detection accuracy and high false alarm rate are caused. Disclosure of Invention The invention provides an intelligent space operation and maintenance optimization method and device based on data elements, which provide an interpretable anomaly analysis path and a comprehensive situation awareness report for operation and maintenance personnel, remarkably improve the fault processing efficiency and realize intelligent operation and maintenance management of an intelligent space. In a first aspect, the present invention provides a data element-based intelligent space operation and maintenance optimization method, which includes: acquiring operation data of an Internet of things terminal, an edge node and a cloud platform in an intelligent space, and fusing the operation data to obtain fused data; sensitive information identification and encryption are carried out on the fusion data, so that encrypted data are obtained; Inputting the encrypted data into local models of all edge nodes, extracting equipment operation characteristics, and integrating equipment operation characteristic sets in a cloud aggregation server; Performing abnormality recognition based on the equipment operation feature set to obtain equipment abnormal state information; And carrying out anomaly cause traceability analysis according to the equipment anomaly state information to obtain an intelligent space anomaly situation report. With reference to the first aspect, in a first implementation manner of the first aspect of the present invention, the collecting operation data of the internet of things terminal, the edge node and the cloud platform in the smart space, and fusing the operation data to obtain fused data includes: Acquiring respective operation data from an internet of things terminal, an edge node and a cloud platform in an intelligent space respectively; Calculating the terminal edge similarity of the terminal data of the Internet of things and the edge node data in the operation data; Calculating edge cloud similarity and equipment time coordination of edge node data and cloud platform data in the operation data; And carrying out association fusion based on the terminal edge similarity, the edge cloud similarity and the equipment time coordination to obtain fusion data. With reference to the first aspect, in a second implementation manner of the first aspect of the present invention, the calculating edge cloud similarity and equipment time coordination of edge node data and cloud platform data in the running data includes: extracting the numerical range, the data type and the change trend characteristics of the edge node data from the operation data, constructing an edge node characteristic vector, extracting corresponding characteristic parameters of cloud platform data, and constructing a cloud platform characteristic vector; Cosine similarity calculation is carried out on the edge node feature vector and the cloud platform feature vector, and edge cloud similarity is obtained; Establishing a sliding time window based on the timestamp information of the operation data, and analyzing the arrival time difference and the data updating frequency of the edge node data and the cloud platform data in the same sliding time window to obtain a time synchronization analysis result; and calculating the time delay variance and the frequency matching degree of the edge node data and the cloud platf