Automatic deployment management method, system and equipment oriented to autonomous controllable operating system and medium

Abstract

The invention discloses an automatic deployment management method, system and device for an autonomous controllable operating system and a medium, and the method comprises the steps: collecting multi-source data, carrying out the preprocessing of the multi-source data, and constructing a safety communication channel based on the preprocessed multi-source data; extracting comprehensive state features of the preprocessed multi-source data, and constructing a joint feature matrix in combination with dynamic key parameters generated by a secure communication channel; calculating the comprehensive weight of each node based on the joint feature matrix, and generating a dynamic weight network topological graph according to the logic connection state of each node; based on the dynamic weight network topological graph, generating a global optimal deployment sequence and calculating dynamic batch parameters, and generating a scheduling strategy in combination with management nodes; and generating an automatic deployment strategy based on the global optimal deployment sequence and the scheduling strategy. According to the invention, the security and adaptability of the deployment process can be improved, collaborative optimization of resource allocation and the network state is realized, and resource competition and communication delay in the deployment process are reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of automated deployment, and in particular to an automated deployment management method, system, device and medium for an autonomous controllable operating system. BACKGROUND

[0002] With the rapid development of information technology, the operating system as the core basic software of computing devices, its autonomous controllable ability has become a key link of national strategic security and industrial development. In recent years, autonomous controllable operating system has been widely used in key fields such as national defense, industrial control, aerospace, and its deployment and management technology has also become a research hotspot. The existing technology proposes a variety of automated deployment solutions, such as solutions based on configuration management tools such as Puppet, Ansible, and microservice architecture solutions combined with container technology and Kubernetes orchestration framework, which to some extent improve the automation level of operating system deployment.

[0003] However, the traditional manual configuration method is not only inefficient, but also prone to errors, and it is difficult to ensure the consistency and stability of deployment; the existing script automation method improves the deployment efficiency to some extent, but lacks flexibility and scalability, and is difficult to adapt to complex and changing deployment environments. In addition, the existing deployment management system usually adopts a monolithic architecture, and the scalability of the management system is poor, which is difficult to meet the needs of large-scale deployment and high concurrency access. SUMMARY

[0004] In view of the above existing problems, the present application is proposed. Therefore, the present application provides an automated deployment management method, system, device and medium for an autonomous controllable operating system to solve the problems of low deployment efficiency and poor scalability of traditional operating systems.

[0005] To solve the above technical problems, the present application provides the following technical solutions:

[0006] In a first aspect, the present application embodiment provides an automated deployment management method for an autonomous controllable operating system, comprising: collecting multi-source data and preprocessing, constructing a secure communication channel based on the preprocessed multi-source data;

[0007] Extracting the comprehensive state features of the preprocessed multi-source data, and combining the dynamic key parameters generated by the secure communication channel to construct a joint feature matrix;

[0008] Based on the joint feature matrix, the comprehensive weight of each node is calculated, and a dynamic weight network topology graph is generated according to the logical connection state of each node;

[0009] Based on the dynamic weight network topology map, a global optimal deployment sequence is generated and dynamic batching parameters are calculated, and a scheduling strategy is generated in combination with a management node;

[0010] Based on the global optimal deployment sequence and the scheduling strategy, an automated deployment strategy is generated.

[0011] As a preferred scheme of the automated deployment management method for the autonomous controllable operating system, wherein: the secure communication channel is constructed based on the preprocessed multi-source data, including:

[0012] Based on the preprocessed multi-source data, a quantum-level entropy source is extracted and a true random number seed is generated;

[0013] Based on the true random number seed, dynamic key parameters are obtained and communication data is encrypted to construct a secure communication channel.

[0014] The preferred technical scheme has the beneficial effects that the secure communication channel generated by the quantum-level entropy source and the global optimal deployment sequence based on the dynamic weight network topology improve the security and adaptability of the deployment process.

[0015] As a preferred scheme of the automated deployment management method for the autonomous controllable operating system, wherein: the comprehensive state features of the preprocessed multi-source data are extracted, and a joint feature matrix is constructed in combination with the dynamic key parameters of the secure communication channel, including:

[0016] The preprocessed multi-source data is arranged row by row according to samples, and the original features corresponding to the samples are aligned column by column to construct a multi-source feature matrix;

[0017] The covariance matrix of the multi-source feature matrix is calculated, the variance contribution rate is obtained and sorted, and the comprehensive state features are extracted;

[0018] Based on the comprehensive state features and the dynamic key parameters of the secure communication channel, a joint feature matrix is constructed.

[0019] As a preferred scheme of the automated deployment management method for the autonomous controllable operating system, wherein: based on the joint feature matrix, the comprehensive weight of each node is calculated, and a dynamic weight network topology map is generated according to the logical connection state of each node, including:

[0020] Based on the joint feature matrix, the weighted Euclidean distance between each node and all other nodes is calculated, and the reciprocal of the average distance from each node to other nodes is taken as the comprehensive weight;

[0021] The comprehensive weight of each node is mapped to the visual attribute of the node, and a dynamic weight network topology map is generated in combination with the logical connection state of each node in the virtual local area network.

[0022] The beneficial effects of this preferred technical solution are that it considers the logical connection between nodes based on virtual local area network (VLAN) partitioning, ensuring that the deployment scheme takes into account the network structure and potential network bottlenecks, enabling better resource allocation and task scheduling, and reducing resource contention and communication latency during the deployment process.

[0023] As a preferred embodiment of the automated deployment and management method for an independently controllable operating system described in this invention, the method involves: generating a globally optimal deployment sequence and calculating dynamic batching parameters based on the dynamic weighted network topology graph, and generating a scheduling strategy in conjunction with the management node, including:

[0024] A fitness function is constructed based on a dynamic weighted network topology graph to generate a globally optimal deployment sequence;

[0025] Dynamic batching parameters are obtained based on preprocessed multi-source data, and management nodes are generated based on the dynamic weighted network topology graph.

[0026] Based on the globally optimal deployment sequence and the real-time status of the management nodes, the task scheduling order and resource allocation ratio are obtained, and a scheduling strategy is generated.

[0027] As a preferred embodiment of the automated deployment management method for an independently controllable operating system described in this invention, the step of generating an automated deployment strategy based on a globally optimal deployment sequence and scheduling strategy includes:

[0028] Based on the preprocessed multi-source data, the lowest latency path between all nodes is calculated, and dynamic reference values ​​reflecting network connectivity and communication efficiency are generated.

[0029] The logical topology and priority order of the nodes to be deployed are obtained based on the globally optimal deployment sequence, and an initial deployment plan is generated in combination with the scheduling strategy.

[0030] Based on the initial deployment plan, collect resource usage and network connection data of the nodes to be deployed, and compare them with dynamic reference values ​​in real time to obtain dynamic early warning signals;

[0031] Based on the dynamic early warning signals, the anomaly type and impact range are identified, and the initial deployment plan is adjusted by combining preprocessed multi-source data and dynamic reference values ​​to generate an automated deployment strategy.

[0032] As a preferred embodiment of the automated deployment and management method for an independently controllable operating system described in this invention, the multi-source data includes: basic device information, network status data, operating indicators, security-related data, and deployment history and maintenance records.

[0033] Secondly, the present invention provides an automated deployment and management system for an independently controllable operating system, comprising:

[0034] The data acquisition module is used to collect multi-source data and preprocess it, and to build a secure communication channel based on the preprocessed multi-source data.

[0035] The secure communication module is used to extract the comprehensive state features of the preprocessed multi-source data and, in combination with the dynamic key parameters generated by the secure communication channel, construct a joint feature matrix.

[0036] The feature extraction and topology generation module is used to calculate the comprehensive weight of each node based on the joint feature matrix and generate a dynamic weighted network topology graph according to the logical connection state of each node.

[0037] The deployment planning module is used to generate a globally optimal deployment sequence and calculate dynamic batching parameters based on the dynamic weighted network topology graph, and generate a scheduling strategy in conjunction with the management node.

[0038] The scheduling strategy generation module is used to generate automated deployment strategies based on the globally optimal deployment sequence and scheduling strategy.

[0039] Thirdly, the present invention provides an electronic device, comprising:

[0040] Memory and processor;

[0041] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the automated deployment and management method for an autonomous and controllable operating system are implemented.

[0042] Fourthly, the present invention provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the automated deployment and management method for an autonomous and controllable operating system.

[0043] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention generates a secure communication channel with dynamic keys through a quantum-level entropy source, and improves the security and adaptability of the deployment process through a globally optimized deployment sequence based on dynamic weighted network topology. Through dynamic weighted topology awareness and batch scheduling strategies under a microservice architecture, it achieves coordinated optimization of resource allocation and network status, reducing resource contention and communication latency during the deployment process. Attached Figure Description

[0044] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0045] Fig. 1 This is a schematic diagram of a method flow for an automated deployment and management method for an independently controllable operating system according to an embodiment of the present invention;

[0046] Fig. 2 This is a schematic diagram of a multi-objective optimization adaptive deployment strategy process for an automated deployment management method for an autonomous and controllable operating system, as described in one embodiment of the present invention.

[0047] Fig. 3 This is a schematic diagram of the dynamic weighted network topology generation and interaction process of an automated deployment and management method for an autonomous and controllable operating system according to an embodiment of the present invention.

[0048] Fig. 4 This is a schematic diagram illustrating the process of generating an autonomous and controllable deployment strategy for an automated deployment management method for an autonomous and controllable operating system, as described in one embodiment of the present invention. Detailed Implementation

[0049] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0050] Example 1, referring to Figs. 1-4 This is one embodiment of the present invention, which provides an automated deployment and management method for an independently controllable operating system, comprising:

[0051] S100: Collects multi-source data and preprocesses it, and builds a secure communication channel based on the preprocessed multi-source data;

[0052] S200: Extract the comprehensive state features of the preprocessed multi-source data and combine them with the dynamic key parameters generated by the secure communication channel to construct a joint feature matrix;

[0053] S300: Calculates the comprehensive weight of each node based on the joint feature matrix, and generates a dynamic weighted network topology based on the logical connection state of each node;

[0054] S400: Based on the dynamic weighted network topology, it generates the globally optimal deployment sequence and calculates dynamic batching parameters, and combines the management node to generate a scheduling strategy.

[0055] S500: Generates automated deployment strategies based on globally optimal deployment sequences and scheduling policies.

[0056] It should be noted that traditional manual configuration methods are not only inefficient but also prone to errors, making it difficult to guarantee deployment consistency and stability. While existing script-based automation methods improve deployment efficiency to some extent, they lack flexibility and scalability, making them unsuitable for complex and ever-changing deployment environments. Furthermore, existing deployment management systems typically employ a monolithic architecture, resulting in poor scalability and difficulty in handling large-scale deployments and high-concurrency access demands. This invention improves the security and adaptability of the deployment process by generating a secure communication channel with dynamic keys using a quantum-level entropy source and a globally optimized deployment sequence based on dynamic weighted network topology. Through dynamic weighted topology awareness and batch scheduling strategies under a microservice architecture, it achieves coordinated optimization of resource allocation and network status, reducing resource contention and communication latency during deployment.

[0057] In this embodiment of the invention, the multi-source data in step S100 includes: basic device information, network status data, operating indicators, security-related data, and deployment history and operation and maintenance records.

[0058] It should be noted that basic equipment information is mainly obtained periodically through configuration management tools such as SNMP protocol, IPMI or cloud platform API to obtain hardware specifications and asset data;

[0059] In one optional embodiment, network status data relies on SNMP probes, NetFlow / sFlow traffic analysis tools, and active Ping / ICMP probing to monitor metrics such as link latency and bandwidth utilization in real time; operational metrics are collected through Prometheus monitoring agents, application instrumentation (such as OpenTelemetry), and log parsing to collect performance data such as CPU / memory usage and service response time; security-related data is obtained from firewall logs and EDR endpoint protection tools (such as CrowdStrike) to obtain attack interception and abnormal behavior records; deployment history and operation and maintenance records are obtained through CI / CD pipelines (such as Jenkins), versioned change logs of configuration management tools, change content, and manual operation records, and are ultimately uniformly connected to a message queue (such as Kafka) for real-time processing and storage.

[0060] In this embodiment of the invention, the preprocessing in step S100 includes data cleaning and format conversion.

[0061] In this embodiment of the invention, the preprocessing in step S100 further includes deduplication, normalization, and outlier handling.

[0062] Furthermore, the collected multi-source data, including basic device information, network status data, operational metrics, security-related data, and deployment history and maintenance records, undergoes data cleaning to remove incomplete, erroneous, or irrelevant data records. Format conversion is performed to ensure all data sources adhere to a unified format standard for subsequent processing. Then, deduplication is performed to eliminate duplicate data entries, ensuring the uniqueness of the dataset. Normalization maps data values ​​from different sources to the same scale, preventing certain features from unduly affecting the results due to significant differences in magnitude. Finally, outlier handling identifies and corrects data points that deviate from the normal range, ensuring the accuracy and reliability of the dataset.

[0063] In this embodiment of the invention, step S100, which involves constructing a secure communication channel based on preprocessed multi-source data, includes:

[0064] Based on the preprocessed multi-source data, quantum-level entropy sources are extracted and true random number seeds are generated.

[0065] Based on a true random number seed, dynamic key parameters are obtained and communication data is encrypted to build a secure communication channel.

[0066] Furthermore, based on the preprocessed multi-source data, high-entropy feature values ​​are first selected from the device basic information and network status data as the initial entropy source through data feature extraction algorithms. Then, the high-entropy feature values ​​are input into the physical layer of the quantum random number generator, such as a photonic quantum random number chip based on single-photon detection or a superposition state measurement device of superconducting qubits. The quantum no-cloning principle is used to amplify the initial entropy source at the quantum level. Then, the quantum superposition state is converted into a classical binary sequence through quantum physical measurements (such as photon arrival time measurement and quantum state collapse observation). At the same time, classical post-processing algorithms are combined to eliminate measurement bias, and finally, a true random number seed that meets cryptographic security standards is generated.

[0067] Furthermore, based on true random number seed initialization, a quantum cryptographic algorithm (such as the lattice-based Kyber algorithm or the hash-based SPHINCS+ algorithm) is used to generate public and private key pairs for both communicating parties through a key generation protocol. Subsequently, the two parties exchange public keys through a secure channel and dynamically calculate session key parameters based on a post-quantum key negotiation protocol. The session key parameters are dynamically updated with each session or time period to ensure forward key security. Then, the dynamically generated session key parameters are used to encrypt communication data (including sensitive data such as task scheduling instructions, node status information, and resource allocation results) (e.g., using the AES-256-GCM symmetric encryption algorithm to protect data confidentiality, while adding an HMAC-SHA3 message authentication code to prevent tampering), ultimately constructing an end-to-end secure communication channel.

[0068] In this embodiment of the invention, step S200 involves extracting the comprehensive state features of the preprocessed multi-source data and combining them with the dynamic key parameters generated by the secure communication channel to construct a joint feature matrix, including:

[0069] The preprocessed multi-source data is arranged row by row according to the samples, and the original features corresponding to the samples are aligned column by column to construct a multi-source feature matrix.

[0070] Calculate the covariance matrix of the multi-source feature matrix, obtain the variance contribution rate and sort them, and extract the comprehensive state features;

[0071] A joint feature matrix is ​​constructed based on comprehensive state characteristics and dynamic key parameters of the secure communication channel.

[0072] Furthermore, the preprocessed multi-source data is traversed row by row, and all the original features (such as CPU utilization, memory usage, network latency, etc.) of each sample (such as monitoring data at each time point) are arranged in a fixed order as columns to form a row of a matrix; this process is repeated until all samples are converted into matrix rows, and finally a multi-source feature matrix is ​​constructed with samples as rows and original features as columns, ensuring that each element corresponds to a specific feature value of a sample, and the matrix dimension is the number of samples × the number of features.

[0073] In a preferred embodiment, the covariance matrix of the multi-source feature matrix is ​​calculated using the PCA algorithm, the variance contribution rate is obtained and sorted, and the comprehensive state features are extracted.

[0074] In an optional embodiment, the variance contribution index of the covariance matrix and the state projection direction vector are calculated using the PCA algorithm. Then, the principal components are sorted from high to low according to their variance contribution rates, and the top N principal components whose cumulative variance contribution rates reach a preset variance contribution rate threshold (determined comprehensively based on data variance preservation requirements, computational efficiency, and task objectives) are selected as key feature directions. Finally, the multi-source data is projected onto the key feature directions to extract the comprehensive state feature vector that can preserve data information to the greatest extent and remove redundant noise.

[0075] It should be noted that PCA (Principal Component Analysis) is a classic dimensionality reduction algorithm. Its core idea is to project the original high-dimensional data into a new coordinate system through orthogonal transformation. The axes (principal components) of this new coordinate system are arranged in order of data variance. The direction of the first principal component is the direction of the largest data variance, and the subsequent principal components are orthogonal and their variances decrease. PCA retains the principal components with high variance contribution rates and discards the components with low variance contribution rates, thereby removing redundancy and noise while preserving data information to the maximum extent. It achieves data dimensionality reduction and feature fusion and is widely used in pattern recognition, anomaly detection and other fields.

[0076] Furthermore, the comprehensive state feature vector obtained after PCA dimensionality reduction is aligned with the dynamic key parameters to ensure that they are strictly synchronized in the time dimension or sampling points. Then, the comprehensive state feature vector and the dynamic key parameter vector are merged into an extended feature vector by horizontal concatenation. The vectors are then stacked according to the time series or sample batches to form a joint feature matrix. Finally, the joint feature matrix is ​​normalized (e.g., Min-Max scaling) to eliminate the influence of different dimensions and ensure that each feature dimension has a balanced contribution during subsequent anomaly detection, thereby constructing the joint feature matrix.

[0077] In this embodiment of the invention, step S300, which calculates the comprehensive weight of each node based on the joint feature matrix and generates a dynamic weighted network topology graph according to the logical connection state of each node, includes:

[0078] Based on the joint feature matrix, the weighted Euclidean distance between each node and all other nodes is calculated, and the inverse of the average distance from each node to other nodes is used as the comprehensive weight.

[0079] The comprehensive weight of each node is mapped to the node's visual attributes, and combined with the logical connection status of each node in the virtual local area network, a dynamic weighted network topology diagram is generated.

[0080] Specifically, the weighted Euclidean distance is calculated as follows:

[0081]

[0082] Where, d ij x represents the weighted Euclidean distance between node i and node j. ik Let x represent the value of the i-th node in the k-th dimension. jk w represents the value of the j-th node in the k-th dimension. k The weight of the k-th dimension is represented by , and m represents the total number of dimensions of the joint feature matrix.

[0083] Link status is obtained from multi-source data using the SNMP protocol and network traffic probes.

[0084] In an optional embodiment, Get-Request messages are periodically sent to network devices via the SNMP protocol to poll their MIB database for VLAN-related OID information (such as ifTable interface status, dot1qVlanCurrentTable VLAN configuration) and obtain basic data such as port UP / DOWN status and VLAN membership. Simultaneously, network traffic probes are deployed to capture data packets within VLANs, parsing their VLAN tags, source / destination IPs, transport layer ports, and traffic rates to obtain VLAN logical connection status. Subsequently, the two types of data are correlated and cleaned. The static configuration information obtained by SNMP and the dynamic traffic characteristics (such as bandwidth utilization and packet loss rate) captured by the traffic probes are timestamped and formatted to eliminate outliers. Finally, the link health status is determined by comprehensively considering port status, VLAN membership validity, and traffic characteristics, and the link status is output, including link ID, current status (normal / congested / interrupted), and key indicators (latency, packet loss rate).

[0085] It should be noted that SNMP (Simple Network Management Protocol) is an application layer protocol used for the management and monitoring of network devices (such as switches, routers, and servers). It transmits management information via the UDP / IP protocol stack and adopts a "manager-agent" architecture. The management end obtains predefined OID (Object Identifier) ​​data from the device agent through polling or trap mechanisms, including standardized information such as device status, interface traffic, and error counts. It supports real-time collection of network topology and performance parameters such as VLAN configuration and port status.

[0086] It should also be noted that a network traffic probe is a dedicated hardware or software tool deployed at critical network nodes (such as mirror ports of core switches and TAP offloading devices). It captures and analyzes data packets from the link layer to the application layer through deep packet inspection technology. It can parse micro-traffic characteristics such as VLAN tags, flow identifiers (such as 5-tuples), and protocol types, and statistically analyze real-time indicators such as link bandwidth utilization, packet drop rate, and latency distribution, and output structured traffic data. It is commonly used for network performance monitoring, security auditing, and fault location.

[0087] In an optional implementation, linear interpolation is used to map the comprehensive weight of each node to the node's visual attributes, and combined with the VLAN logical connection status, the node's visual attributes are read by a graphics rendering engine to generate a dynamic weighted network topology.

[0088] Specifically, a linear interpolation algorithm is used to map the comprehensive weight value of each node to a preset range of visual attributes based on visual recognition requirements (e.g., node size range [5px, 50px], color gradient range [green #00FF00 → red #FF0000]). The specific attribute value corresponding to each node is calculated (e.g., node size = 5 + (current weight / maximum weight) × 45, RGB color values ​​are dynamically generated through linear interpolation). Then, the topological connection relationship between nodes is determined based on the VLAN logical connection status (e.g., generating connections between nodes according to VLAN port interconnection rules, or removing invalid connections according to ACL rules). Finally, the node ID and its mapped visual attributes are compared with the node connections determined based on the VLAN status. The connections are serialized into a dynamic data file. After the graphics rendering engine (such as D3.js, ECharts, or Cytoscape.js) reads the data file, it parses the attribute parameters of nodes and edges, calls the Canvas or WebGL graphics interface to draw the basic network topology graph, and applies the force-directed layout algorithm to automatically adjust the distribution of nodes. Then, it obtains the latest weight data through timed polling (such as requesting the backend API every 3 seconds) or WebSocket real-time push mechanism, dynamically updates the visual attributes of nodes and edges, triggers the rendering engine to redraw locally or globally, and finally generates a dynamic weighted network topology graph that can reflect the changes in network status in real time, intuitively presenting the distribution of high-load nodes and critical links.

[0089] In this embodiment of the invention, step S400, based on the dynamic weighted network topology graph, generates a globally optimal deployment sequence and calculates dynamic batching parameters, and combines this with the management node to generate a scheduling strategy, including:

[0090] A fitness function is constructed based on a dynamic weighted network topology graph to generate a globally optimal deployment sequence;

[0091] Dynamic batching parameters are obtained based on preprocessed multi-source data, and management nodes are generated based on the dynamic weighted network topology graph.

[0092] Based on the globally optimal deployment sequence and the real-time status of the management nodes, the task scheduling order and resource allocation ratio are obtained, and a scheduling strategy is generated.

[0093] In this embodiment of the invention, the fitness function expression in step S400 is:

[0094] f=α∑ n q n r n -β∑ (n,p)∈E c np y np ;

[0095] Where α represents the weight coefficient of the node, with a value in the range α≥0, and β represents the weight coefficient of the link, with a value in the range β≥0, and satisfying α+β=1, q n r represents the overall weight of node n. n c represents the deployment state variable of node n. np y represents the total load of link (n,p). np The variable represents the usage of link (n, p), where E represents the set of links, p represents the terminating node of the link, and n represents the index variable of the node in the dynamic weighted network topology graph.

[0096] In an optional embodiment, step S400, which generates a globally optimal deployment sequence and calculates dynamic batching parameters based on the dynamic weighted network topology graph, and combines the management node to generate a scheduling strategy, further includes:

[0097] The K-means clustering algorithm is used to group the original features of the preprocessed multi-source data. The task batches are automatically divided according to the similarity of task resource requirements and node load characteristics. The batching parameters (such as the composition of each batch of tasks and the resource allocation ratio) are dynamically determined. At the same time, in the dynamic weighted network topology, each node broadcasts and compares weight information to each other through a distributed election protocol based on real-time calculated weight indicators such as load and communication latency. Finally, the node with the highest comprehensive weight and network stability is elected as the management node, which coordinates the global task scheduling and resource allocation.

[0098] In another optional embodiment, step S400, based on the dynamic weighted network topology graph, generates a globally optimal deployment sequence and calculates dynamic batching parameters, and combines this with the management node to generate a scheduling strategy, further includes:

[0099] The DBSCAN clustering algorithm is used to group the original features of the preprocessed multi-source data. The task batches are automatically divided according to the similarity of task resource requirements and node load characteristics. The batching parameters (such as the composition of each batch of tasks and the resource allocation ratio) are dynamically determined. At the same time, in the dynamic weighted network topology, each node broadcasts and compares weight information to each other through a distributed election protocol based on real-time calculated weight indicators such as load and communication latency. Finally, the node with the highest comprehensive weight and the most stable network is elected as the management node, which coordinates the global task scheduling and resource allocation.

[0100] It should be noted that clustering algorithms are a type of unsupervised learning algorithm used to automatically divide highly similar samples in a dataset into different groups without pre-labeling the categories. Its core idea is to iteratively optimize intra-cluster compactness and inter-cluster separation by calculating the distance or similarity between samples. Common clustering algorithms include K-means (partition-based) and DBSCAN (density-based). In dynamic batching scenarios, clustering algorithms automatically group tasks into batches based on the similarity of task resource requirements (such as CPU and memory usage) and node status (such as remaining computing power and load), thereby maximizing resource utilization and load balancing.

[0101] Furthermore, distributed election protocols are communication protocols used to autonomously negotiate and elect a unique leader (management node) among multiple nodes, and must satisfy fault tolerance, consistency, and efficiency. Its core process includes nodes broadcasting their own state (such as weight and availability), determining the leader through mechanisms such as voting, comparison, or token passing, and re-triggering an election when the leader fails. Common protocols include Raft (based on log replication and heartbeat detection) and the Bully algorithm (based on priority comparison). In dynamic weighted network topologies, nodes participate in the election based on real-time calculated weight indicators such as load and bandwidth, ensuring that the elected management node has the optimal global perspective and communication capabilities, thereby reliably coordinating task scheduling and resource allocation.

[0102] In an optional embodiment, step S400, which generates a globally optimal deployment sequence and calculates dynamic batching parameters based on the dynamic weighted network topology graph, and combines the management node to generate a scheduling strategy, further includes:

[0103] By combining the globally optimal deployment sequence with the real-time status of management nodes, a greedy algorithm is used to obtain the task scheduling order and resource allocation ratio, and a scheduling strategy is generated.

[0104] Specifically, based on the globally optimal deployment sequence and dynamic batching parameters, the system first selects a set of available nodes that meet the resource requirements of the current batch of tasks based on the real-time status of the management nodes (including CPU utilization, memory usage, network bandwidth load, etc.). Then, according to the task priority order of the globally optimal deployment sequence, the current batch of tasks is allocated one by one to the available nodes with the most abundant remaining resources and the lowest link cost, while updating the node resource occupancy status and link usage flags. Next, the batching parameters for the next batch are dynamically adjusted based on the real-time monitored changes in node load. After the current batch of tasks is allocated, the status of the management nodes is re-collected to verify the feasibility of scheduling the next batch. Finally, a scheduling scheme is generated that includes the dynamic batching task allocation results, the resource usage ratio of the nodes to be deployed, and the link selection strategy.

[0105] It should be noted that a greedy algorithm is an algorithmic strategy that adopts the optimal decision in the current state at each step. It approaches the global optimal solution by gradually accumulating local optimal solutions. Its core feature is that it only focuses on the optimal choice in the current step without backtracking to adjust previous decisions. It is suitable for problem scenarios with greedy selection properties and optimal substructure characteristics, such as prioritizing the task with the highest resource demand in task scheduling to quickly reduce the overall load, or selecting the node closest to the target at each step in path planning to shorten the total path length.

[0106] In this embodiment of the invention, step S500, which generates an automated deployment strategy based on the globally optimal deployment sequence and scheduling strategy, includes:

[0107] Based on the preprocessed multi-source data, the lowest latency path between all nodes is calculated, and dynamic reference values ​​reflecting network connectivity and communication efficiency are generated.

[0108] The logical topology and priority order of the nodes to be deployed are obtained based on the globally optimal deployment sequence, and an initial deployment plan is generated in combination with the scheduling strategy.

[0109] Based on the initial deployment plan, collect resource usage and network connection data of the nodes to be deployed, and compare them with dynamic reference values ​​in real time to obtain dynamic early warning signals;

[0110] Based on the dynamic early warning signals, the anomaly type and impact range are identified, and the initial deployment plan is adjusted by combining preprocessed multi-source data and dynamic reference values ​​to generate an automated deployment strategy.

[0111] In an optional embodiment, step S500, which generates an automated deployment strategy based on the globally optimal deployment sequence and scheduling strategy, further includes: extracting the network topology (such as the physical connection relationship of devices such as switches and routers) from the device basic information; obtaining real-time latency data of each link (such as ICMP Ping round-trip time, TCP connection latency, etc.) from network status data and operation indicators; filtering out abnormal links with attacks or faults by combining security-related data; supplementing the stability information of the links from the deployment history and operation and maintenance records; then constructing a weighted directed graph (nodes are network devices, and edge weights are comprehensive latency values, including real-time latency and historical stability correction factors); running Dijkstra's algorithm to calculate the lowest latency path to all other nodes starting from each node (the priority queue dynamically selects the current shortest path for expansion); and finally generating a dynamic routing table containing the optimal path between all node pairs and the corresponding cumulative latency values ​​as a reference for reflecting network connectivity and communication efficiency.

[0112] It should be noted that Dijkstra's shortest path algorithm is a classic algorithm for calculating the shortest path from a single source node to all other nodes in a graph. Its core idea is to gradually expand the set of known shortest paths using a greedy strategy. Starting from the source node, each time the nearest unvisited neighbor node is selected and added to the set, and the path distances of other unvisited nodes are updated based on this neighbor. This process is repeated until the shortest paths of all nodes are determined. It is suitable for weighted directed or undirected graphs with non-negative edge weights and can efficiently solve shortest path problems in scenarios such as network routing optimization and traffic path planning.

[0113] In an optional embodiment, the step S500 of generating an automated deployment strategy based on the globally optimal deployment sequence and scheduling strategy further includes: obtaining the logical topology relationship and priority order of the nodes to be deployed through the NSGA-II algorithm based on the globally optimal deployment sequence, and generating an initial deployment scheme in combination with the scheduling strategy;

[0114] Specifically, based on the globally optimal deployment sequence, the NSGA-II algorithm is used to perform multi-objective optimization of tasks and nodes, with the shortest task completion time and the highest resource utilization as optimization objectives. Individuals in the population are divided into different non-dominated levels through non-dominated sorting, and the crowding distance between individuals is calculated to maintain population diversity. Then, in each generation of evolution, genetic operations such as selection, crossover, and mutation are used to update individuals. After multiple iterations, the solution that satisfies the task constraints and has the best overall performance is selected from the final non-dominated solution set, thereby determining the logical topology relationship (such as the dependency connection and communication path between tasks) and priority ranking (based on the criticality of tasks and the urgency of resource requirements) of the nodes to be deployed. Finally, the logical topology relationship and priority ranking obtained above are combined with the established scheduling strategy, taking into account factors such as the real-time load of nodes, communication bandwidth, and resource quotas allocated by management nodes, to generate an initial deployment scheme that includes task deployment location, execution order, and initial resource allocation ratio.

[0115] It should be noted that the NSGA-II algorithm is a classic multi-objective optimization algorithm, primarily used to solve optimization problems with multiple conflicting objectives. Based on a genetic algorithm framework, the NSGA-II algorithm divides individuals in the population into different frontiers through non-dominated sorting, grouping individuals without dominance relationships into one category to determine the hierarchy of individual merit. Simultaneously, it calculates individual crowding to measure the distribution density of solutions in the objective space, ensuring population diversity. Genetic operations such as tournament selection, simulated binary crossover, and polynomial mutation are used to generate new individuals. An elite retention strategy preserves high-quality individuals from both parents and offspring, preventing the loss of excellent solutions. This process is iterated continuously, gradually approximating the Pareto optimal solution set, ultimately yielding a set of non-dominated solutions that are relatively excellent across multiple objectives, providing decision-makers with multiple alternatives.

[0116] Furthermore, based on the initial deployment plan, the scope of monitoring objects is first determined by basic device information (such as device types and identifiers like servers, switches, and routers). Real-time resource metrics such as CPU utilization, memory usage, and disk read / write speeds for each node are collected from network status data. Simultaneously, connection status data such as network link latency, bandwidth utilization, and packet loss rate are acquired. Then, historical load fluctuation patterns are extracted from operational metrics, and combined with security-related data, abnormal events (such as port scanning, traffic surges, and other security threats) and failure modes (such as high-frequency downtime records of specific nodes) are identified from deployment history and operation and maintenance records. Then, the real-time collected resource and network data are compared with dynamic reference values ​​in multiple dimensions. In terms of resources, the comparison is made to see if CPU, memory, etc. have exceeded the historical fluctuation limit or safety baseline. In terms of network, the comparison is made to see if the latency is higher than the historical best value. At the same time, the abnormal events and failure modes are combined to determine whether there are potential risks. Finally, the abnormal indicators are judged in real time, and dynamic early warning signals are generated, including node identifier, resource / network type, abnormal value, deviation degree and timestamp (such as "Node 5 memory usage reaches 92%, exceeding the historical peak of 85%" or "Link CD latency suddenly increased by 150ms and DDoS attack was detected").

[0117] Furthermore, the anomaly self-healing mechanism first performs real-time analysis of dynamic early warning signals to identify the anomaly type (such as CPU overload, memory leak, network latency spike, or link interruption) and its impact range (affected nodes, links, or services). At the same time, it comprehensively assesses the severity of the anomaly by combining preprocessed multi-source data and dynamic reference values. Then, it dynamically adjusts the initial deployment plan based on the assessment results. For nodes with insufficient computing resources, it selects idle nodes from the resource pool and sorts them according to their remaining CPU / memory. It prioritizes selecting the node with the lowest load and the lowest network latency as the expansion target. For network failure links, it recalculates the lowest latency path based on the real-time network topology and switches the affected traffic to the optimal backup path, ultimately generating an automated deployment strategy.

[0118] Example 2: The above example is an illustrative scheme of an automated deployment management method for an independently controllable operating system. It should be noted that the technical solution of this automated deployment management system for an independently controllable operating system belongs to the same concept as the technical solution of the above-described automated deployment management method for an independently controllable operating system. Details not described in detail in this example of the automated deployment management system for an independently controllable operating system can be found in the description of the above-described automated deployment management method for an independently controllable operating system.

[0119] This embodiment describes an automated deployment management system for an independently controllable operating system, comprising:

[0120] The data acquisition module is used to collect multi-source data and preprocess it, and to build a secure communication channel based on the preprocessed multi-source data.

[0121] The secure communication module is used to extract the comprehensive state features of the preprocessed multi-source data and, in combination with the dynamic key parameters generated by the secure communication channel, construct a joint feature matrix.

[0122] The feature extraction and topology generation module is used to calculate the comprehensive weight of each node based on the joint feature matrix and generate a dynamic weighted network topology graph according to the logical connection state of each node.

[0123] The deployment planning module is used to generate the globally optimal deployment sequence and calculate dynamic batching parameters based on the dynamic weighted network topology graph, and generate a scheduling strategy in conjunction with the management node.

[0124] The scheduling strategy generation module is used to generate automated deployment strategies based on the globally optimal deployment sequence and scheduling strategy.

[0125] This embodiment also provides an electronic device applicable to automated deployment and management methods for autonomous and controllable operating systems, including:

[0126] The system includes a memory and a processor. The memory stores computer-executable instructions, and the processor executes these instructions to implement the automated deployment and management method for an autonomous and controllable operating system as proposed in the above embodiments.

[0127] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, it implements the automated deployment and management method for an autonomous and controllable operating system as proposed in the above embodiments.

[0128] The storage medium proposed in this embodiment and the automated deployment and management method for an autonomous and controllable operating system proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0129] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0130] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. An automated deployment and management method for an independently controllable operating system, characterized in that, include: Collect and preprocess multi-source data, and build a secure communication channel based on the preprocessed multi-source data; Extract the comprehensive state features of the preprocessed multi-source data and combine them with the dynamic key parameters generated by the secure communication channel to construct a joint feature matrix; The comprehensive weight of each node is calculated based on the joint feature matrix, and a dynamic weighted network topology is generated according to the logical connection state of each node. Based on the dynamic weighted network topology, a globally optimal deployment sequence is generated and dynamic batching parameters are calculated. A scheduling strategy is then generated in conjunction with the management node. An automated deployment strategy is generated based on the globally optimal deployment sequence and scheduling strategy.

2. The automated deployment and management method for an independently controllable operating system as described in claim 1, characterized in that, The construction of a secure communication channel based on preprocessed multi-source data includes: Based on the preprocessed multi-source data, quantum-level entropy sources are extracted and true random number seeds are generated. Based on a true random number seed, dynamic key parameters are obtained and communication data is encrypted to build a secure communication channel.

3. The automated deployment and management method for an independently controllable operating system as described in claim 2, characterized in that, Extracting the comprehensive state features of the preprocessed multi-source data and combining them with the dynamic key parameters generated by the secure communication channel, a joint feature matrix is ​​constructed, including: The preprocessed multi-source data is arranged row by row according to the samples, and the original features corresponding to the samples are aligned column by column to construct a multi-source feature matrix. Calculate the covariance matrix of the multi-source feature matrix, obtain the variance contribution rate and sort them, and extract the comprehensive state features; A joint feature matrix is ​​constructed based on comprehensive state characteristics and dynamic key parameters of the secure communication channel.

4. The automated deployment and management method for an independently controllable operating system as described in claim 3, characterized in that, The comprehensive weight of each node is calculated based on the joint feature matrix, and a dynamic weighted network topology graph is generated according to the logical connection state of each node, including: Based on the joint feature matrix, the weighted Euclidean distance between each node and all other nodes is calculated, and the inverse of the average distance from each node to other nodes is used as the comprehensive weight. The comprehensive weight of each node is mapped to the node's visual attributes, and combined with the logical connection status of each node in the virtual local area network, a dynamic weighted network topology diagram is generated.

5. The automated deployment and management method for an independently controllable operating system as described in claim 4, characterized in that, Based on the dynamic weighted network topology, a globally optimal deployment sequence is generated and dynamic batching parameters are calculated. A scheduling strategy is then generated in conjunction with the management node, including: A fitness function is constructed based on a dynamic weighted network topology graph to generate a globally optimal deployment sequence; Dynamic batching parameters are obtained based on preprocessed multi-source data, and management nodes are generated based on the dynamic weighted network topology graph. Based on the globally optimal deployment sequence and the real-time status of the management nodes, the task scheduling order and resource allocation ratio are obtained, and a scheduling strategy is generated.

6. The automated deployment and management method for an independently controllable operating system as described in claim 5, characterized in that, The generation of automated deployment strategies based on globally optimal deployment sequences and scheduling strategies includes: Based on the preprocessed multi-source data, the lowest latency path between all nodes is calculated, and dynamic reference values ​​reflecting network connectivity and communication efficiency are generated. The logical topology and priority order of the nodes to be deployed are obtained based on the globally optimal deployment sequence, and an initial deployment plan is generated in combination with the scheduling strategy. Based on the initial deployment plan, collect resource usage and network connection data of the nodes to be deployed, and compare them with dynamic reference values ​​in real time to obtain dynamic early warning signals; Based on the dynamic early warning signals, the anomaly type and impact range are identified, and the initial deployment plan is adjusted by combining preprocessed multi-source data and dynamic reference values ​​to generate an automated deployment strategy.

7. The automated deployment and management method for an independently controllable operating system as described in claim 6, characterized in that, Multi-source data includes: basic device information, network status data, operational metrics, security-related data, and deployment history and maintenance records.

8. An automated deployment management system for an independently controllable operating system, applied to the method described in any one of claims 1-7, characterized in that, include: The data acquisition module is used to collect multi-source data and preprocess it, and to build a secure communication channel based on the preprocessed multi-source data. The secure communication module is used to extract the comprehensive state features of the preprocessed multi-source data and, in combination with the dynamic key parameters generated by the secure communication channel, construct a joint feature matrix. The feature extraction and topology generation module is used to calculate the comprehensive weight of each node based on the joint feature matrix and generate a dynamic weighted network topology graph according to the logical connection state of each node. The deployment planning module is used to generate a globally optimal deployment sequence and calculate dynamic batching parameters based on the dynamic weighted network topology graph, and generate a scheduling strategy in conjunction with the management node. The scheduling strategy generation module is used to generate automated deployment strategies based on the globally optimal deployment sequence and scheduling strategy.

9. An electronic device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, they implement the steps of the automated deployment and management method for an autonomous and controllable operating system as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the automated deployment and management method for an autonomous and controllable operating system as described in any one of claims 1 to 7.

Images