CN111612300A - A scene anomaly perception index calculation method and system based on deep hybrid cloud model - Google Patents

Abstract

The invention relates to the technical field of index calculation, in particular to a method and system for calculating a scene abnormality perception index based on a deep hybrid cloud model. The method includes the following steps: using graphembedding to perform graph coding on network resource nodes; using deep learning-based analytic hierarchy process to establish a cloud model evaluation scale; using the cloud scale to perform online scene anomaly perception detection. In the method and system for calculating scene anomaly perception indicators based on the deep hybrid cloud model, graph embedding is used to represent network resource nodes with graph coding, which has a wider scope of application and generalization ability, and the coded node resources are used to construct an overall cloud model. , and use the deep learning-based analytic hierarchy process to establish a cloud model evaluation scale for scene anomaly perception detection, which can be detected under data of different periodic dimensions, the training model is used for online detection, and can be used for large-scale, dynamic changes. Real-time detection of network resources.

Description

A kind of scene anomaly perception index calculation method based on deep hybrid cloud model and system

technical field

The invention relates to the technical field of index calculation, in particular to a method and system for calculating a scene abnormality perception index based on a deep hybrid cloud model.

Background technique

Existing detection systems cannot adapt to large-scale and dynamically changing network resources, and can only sense the presence of abnormal points in the overall network, but cannot precisely locate the abnormal points.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a method and system for calculating a scene abnormality perception index based on a deep hybrid cloud model, so as to solve the problems raised in the above background art.

In order to solve the above technical problems, one of the purposes of the present invention is to provide a method for calculating a scene abnormality perception index based on a deep hybrid cloud model, the method comprising the following steps:

S1. Use graphembedding to perform graph coding on network resource nodes;

S2. Use the deep learning-based analytic hierarchy process to establish a cloud model evaluation scale;

S3. Use the cloud ruler to perform online scene anomaly perception detection.

Preferably, in the S1, the method of using graphembedding to perform graph coding on network resource nodes specifically includes the following steps:

S1.1. Input: Graph (nodes in the graph represent network resources, such as servers, etc., and the nodes (network resources) in each graph contain their own attributes, such as storage capacity, occupancy rate, bandwidth, etc.; The weight of the edge (edge) between nodes can reflect the direct or indirect dependence of each other's network resources);

S1.2. Output: GraphEmbedding-based graph coding representation for network resources;

S1.3. Initialization: adjacency matrix based on network resource (node), feature matrix based on network resource (node) attribute;

S1.4, iterative training GraphEmbedding.

Preferably, in S1.4, the method for iteratively training GraphEmbedding specifically includes the following steps:

S1.4.1. For the graph (Graph) formed by the network resource (node), first analyze the spatial structure of the Graph, and use the Aggregate Function (AggregateFunction) to aggregate the low-order and high-order spatial relationships of any node (node) to form Representation of new spatial relationships for nodes;

S1.4.2. For the aggregated spatial relationship formed in S1.4.1, aggregate the characteristic relationship of the node (node) known in the initialization stage, and integrate the spatial relationship of the network resource (node) through the IoT-GraphEmbedding algorithm. ) and feature relationships;

S1.4.3. Through multiple trainings, optimize the network structure to form the optimal graph coding representation.

Preferably, in the S2, the method for establishing a cloud model evaluation scale using the deep learning-based analytic hierarchy process specifically includes the following steps:

S2.1. The scales of 0.1-0.9 correspond to 9 cloud models (Exi, Eni, Hei) (i=1, 2, ..., 9), where 0.1, 0.2, ..., 0.9 correspond to the cloud model's Expect Ex1, Ex2, ..., Ex9;

S2.2. Suppose the universe of discourse U of the nine cloud models is [0.1, 0.9], and the expected values of each cloud model are Ex ₁ = 0.1, Ex ₂ = 0.2, ..., Ex ₉ = 0.9, according to the golden section method to obtain each The entropy and super-entropy of cloud models, wherein, the entropy of each cloud model: En ₁ =En ₃ =En ₅ =En ₇ =En ₉ =0.0707, En ₂ =En ₄ =En ₆ =En ₈ =0.0437; The superentropy of : He ₁ =He ₃ =He ₅ =He7=He9=0.0118, He2=He4=He6=He8=0.0073. From this, the scale (Exi, En i , He i ) (Exi, En _i , He _i ) ( i = 1, 2, ..., 9);

S2.3. Constructing a judgment matrix, it is assumed that there are m adjacent base clouds C1=(Ex ₁ , En ₁ , He ₁ ), ..., Cm=(Ex _m , in the universe of discourse [U _min , U _max ], En _m , He _m ), the floating cloud C=(Ex, En, He) of qualitative concept can be obtained by gathering m clouds, where the numerical index calculation formula is:

The second purpose of the present invention is to provide a scene abnormality perception index calculation system based on a deep hybrid cloud model, including:

The graph coding module is used to use graphembedding for graph coding of network resource nodes;

Establish a cloud model module for establishing a cloud model evaluation scale using AHP based on deep learning;

The perception detection module is used for online scene anomaly perception detection using the cloud ruler.

The third object of the present invention is to provide a scene abnormality perception index calculation device based on a deep hybrid cloud model, including a processor, a memory, and a computer program stored in the memory and running on the processor, the When the processor executes the above-mentioned computer program, any one of the above-mentioned steps of the calculation method for the scene abnormality perception index based on the deep hybrid cloud model is implemented.

The fourth object of the present invention is to provide a computer-readable storage medium, in which at least one program is stored, and the at least one program is executed by the above-mentioned processor to realize the depth-based method described in any one of the above. The steps of the calculation method of the scene abnormality perception index of the hybrid cloud model.

Compared with the prior art, the present invention has the beneficial effects: in the method and system for calculating the scene abnormality perception index based on the deep hybrid cloud model, graph embedding is used to express the network resource node by graph coding, which is similar to the calculation of the traditional numerical index. More types of network resource nodes can be included in the analysis, such as servers, databases, processes, electricity meters, bandwidth, etc., for indiscriminate analysis, with a wider scope of application and generalization capabilities, and the coded node resources Build an overall cloud model, and use the deep learning-based AHP to establish a cloud model evaluation scale for scene anomaly perception detection, which can be detected under data of different periodic dimensions. Compared with traditional periodic detection, it can meet more requirements usage requirements. The training model is used for online detection, and it can perform real-time detection of large-scale and dynamically changing network resources to meet most actual production environments.

Description of drawings

Fig. 1 is the overall method flow chart of the present invention;

FIG. 2 is a flow chart of a method for graph coding a network resource node using graphembedding according to the present invention;

Fig. 3 is the iterative training GraphEmbedding method flow chart of the present invention;

4 is a flowchart of a method for establishing a cloud model evaluation scale using deep learning-based analytic hierarchy process of the present invention;

FIG. 5 is a block diagram of a scene abnormality perception index calculation system based on a deep hybrid cloud model of the present invention;

FIG. 6 is a structural diagram of an index calculation apparatus according to an embodiment of the present invention.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

Please refer to Fig. 1-Fig. 6, the present invention provides a technical solution:

The present invention provides a method for calculating a scene abnormality perception index based on a deep hybrid cloud model, the method comprising the following steps:

S1. Use graphembedding to perform graph coding on network resource nodes;

S2. Use the deep learning-based analytic hierarchy process to establish a cloud model evaluation scale;

S3. Use the cloud ruler to perform online scene anomaly perception detection.

In this embodiment, in S1, the method of using graphembedding to perform graph coding on network resource nodes specifically includes the following steps:

S1.1. Input: Graph (nodes in the graph represent network resources, such as servers, etc., and the nodes (network resources) in each graph contain their own attributes, such as storage capacity, occupancy rate, bandwidth, etc.; The weight of the edge (edge) between nodes can reflect the direct or indirect dependence of each other's network resources);

S1.2. Output: GraphEmbedding-based graph coding representation for network resources;

S1.3. Initialization: adjacency matrix based on network resource (node), feature matrix based on network resource (node) attribute;

S1.4, iterative training GraphEmbedding.

Further, in S1.4, the method of iteratively training GraphEmbedding specifically includes the following steps:

S1.4.1. For the graph (Graph) formed by the network resource (node), first analyze the spatial structure of the Graph, and use the Aggregate Function (AggregateFunction) to aggregate the low-order and high-order spatial relationships of any node (node) to form Representation of new spatial relationships for nodes;

S1.4.2. For the aggregated spatial relationship formed in S1.4.1, aggregate the characteristic relationship of the node (node) known in the initialization stage, and integrate the spatial relationship of the network resource (node) through the IoT-GraphEmbedding algorithm. ) and feature relationships;

S1.4.3. Through multiple trainings, optimize the network structure to form the optimal graph coding representation.

Specifically, in S2, the method for establishing a cloud model evaluation scale using the deep learning-based AHP method specifically includes the following steps:

S2.1. The scales of 0.1-0.9 correspond to 9 cloud models (Exi, Eni, Hei) (i=1, 2, ..., 9), where 0.1, 0.2, ..., 0.9 correspond to the cloud model's Expect Ex1, Ex2, ..., Ex9;

S2.2. Suppose the universe of discourse U of the nine cloud models is [0.1, 0.9], and the expected values of each cloud model are Ex ₁ = 0.1, Ex ₂ = 0.2, ..., Ex ₉ = 0.9, according to the golden section method to obtain each The entropy and super-entropy of cloud models, wherein, the entropy of each cloud model: En ₁ =En ₃ =En ₅ =En ₇ =En ₉ =0.0707, En ₂ =En ₄ =En ₆ =En ₈ =0.0437; The superentropy of : He ₁ =He ₃ =He ₅ =He7=He9=0.0118, He2=He4=He6=He8=0.0073, from which the scale (Exi, En i , He i ) (Exi, En _i , He _i ) ( i = 1, 2, ..., 9);

S2.3. Constructing a judgment matrix, it is assumed that there are m adjacent base clouds C1=(Ex ₁ , En ₁ , He ₁ ), ..., Cm=(Ex _m , in the universe of discourse [U _min , U _max ], En _m , He _m ), the floating cloud C=(Ex, En, He) of qualitative concept can be obtained by gathering m clouds, where the numerical index calculation formula is:

The second purpose of the present invention is to provide a scene abnormality perception index calculation system based on a deep hybrid cloud model, including:

The graph coding module is used to use graphembedding for graph coding of network resource nodes;

Establish a cloud model module for establishing a cloud model evaluation scale using AHP based on deep learning;

The perception detection module is used for online scene anomaly perception detection using the cloud ruler.

It should be noted that, for the functions of the graph encoding module, the cloud model building module, and the perception detection module, please refer to the description of the method corresponding to each module, and will not be repeated here.

Referring to FIG. 6 , a schematic structural diagram of an apparatus for providing a scene abnormality perception indicator based on a deep hybrid cloud model according to an embodiment of the present invention is shown. The apparatus includes a processor 101 , a memory 102 and a bus 103 .

The processor package 101 includes one or more processing cores, the processor 101 is connected to the processor 101 through a bus 103, the memory 102 is used to store program instructions, and when the processor 102 executes the program instructions in the memory 102, the above-mentioned deep hybrid cloud is implemented The model's scene anomaly perception index calculation method.

Alternatively, memory 102 may be implemented by any type of volatile or non-volatile storage device or combination thereof, such as static anytime access memory (SRAM), electrically erasable programmable read only memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Read Only Memory (PROM), Read Only Memory (ROM), Magnetic Memory, Flash Memory, Magnetic or Optical Disk.

In addition, the present invention also provides a computer-readable storage medium, in which at least one program is stored, and at least one program is executed by the above-mentioned processor to realize the scene anomaly perception indicator based on the deep hybrid cloud model as described above. The steps of the calculation method.

Optionally, the present invention also provides a computer program product including instructions, which, when running on a computer, causes the computer to execute the steps of the above aspects of the method for calculating scene anomaly perception indicators based on the deep hybrid cloud model.

Those of ordinary skill in the art can understand that all or part of the steps of implementing the above embodiments can be completed by hardware, or can be completed by instructing relevant hardware through a program, and the program can be stored in a computer-readable storage medium. The storage medium mentioned may be a read-only memory, a magnetic disk or an optical disk, etc.

The foregoing has shown and described the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above-mentioned embodiments, and the above-mentioned embodiments and descriptions are only preferred examples of the present invention, and are not intended to limit the present invention, without departing from the spirit and scope of the present invention. Under the premise, the present invention will also have various changes and improvements, and these changes and improvements all fall within the scope of the claimed invention. The claimed scope of the present invention is defined by the appended claims and their equivalents.

Claims (7)

1. A method for calculating a scene abnormality perception index based on a deep hybrid cloud model, the method comprising the following steps: S1. Use graphembedding to perform graph coding on network resource nodes; S2. Use the deep learning-based analytic hierarchy process to establish a cloud model evaluation scale; S3. Use the cloud ruler to perform online scene anomaly perception detection. 2. The method for calculating a scene abnormality perception index based on a deep hybrid cloud model according to claim 1, characterized in that: in the S1, the method for performing graph coding on network resource nodes using graphembedding specifically comprises the following steps: S1.1. Input: Graph represents network resources, and each node in the graph contains its own attributes; S1.2. Output: GraphEmbedding-based graph coding representation for network resources; S1.3, initialization: adjacency matrix based on network resources, feature matrix based on network resource attributes; S1.4, iterative training GraphEmbedding. 3. The method for calculating a scene abnormality perception index based on a deep hybrid cloud model according to claim 2, characterized in that: in the S1.4, the method for iteratively training GraphEmbedding specifically comprises the following steps: S1.4.1. For the graph formed by network resources, first analyze the spatial structure of the Graph, and aggregate the low-order and high-order spatial relationships of any node through the aggregation function to form the representation of the new spatial relationship of the node; S1.4.2. For the aggregated spatial relationship formed in S1.4.1, aggregate the feature relationship of nodes known in the initialization stage, and integrate the spatial relationship and feature relationship of network resources through IoT-GraphEmbedding algorithm; S1.4.3. Through multiple trainings, optimize the network structure to form the optimal graph coding representation. 4. The method for calculating a scene abnormality perception index based on a deep hybrid cloud model according to claim 1, wherein in the S2, the method for establishing a cloud model evaluation scale using the deep learning-based analytic hierarchy process specifically comprises the following steps : S2.1. The scales of 0.1-0.9 correspond to 9 cloud models (Exi, Eni, Hei) (i=1, 2, ..., 9), where 0.1, 0.2, ..., 0.9 correspond to the cloud model's Expect Ex1, Ex2, ..., Ex9; S2.2. Suppose the universe of discourse U of the nine cloud models is [0.1, 0.9], and the expected values of each cloud model are Ex ₁ = 0.1, Ex ₂ = 0.2, ..., Ex ₉ = 0.9, according to the golden section method to obtain each The entropy and super-entropy of cloud models, wherein, the entropy of each cloud model: En ₁ =En ₃ =En ₅ =En ₇ =En ₉ =0.0707, En ₂ =En ₄ =En ₆ =En ₈ =0.0437; The superentropy of: He ₁ =He ₃ =He ₅ =He7=He9=0.0118, He2=He4=He6=He8=0.0073; get the scale represented by the cloud model (Exi, En _i , He _i ) (i=1 , 2, ..., 9); S2.3. Constructing a judgment matrix, it is assumed that there are m adjacent base clouds C1=(Ex ₁ , En ₁ , He ₁ ), ..., Cm=(Ex _m , in the universe of discourse [U _min , U _max ], En _m , He _m ), the floating cloud C=(Ex, En, He) of qualitative concept can be obtained by gathering m clouds, where the numerical index calculation formula is:

5. A scene anomaly perception index calculation system based on a deep hybrid cloud model, comprising: The graph coding module is used to use graphembedding for graph coding of network resource nodes; Establish a cloud model module for establishing a cloud model evaluation scale using AHP based on deep learning; The perception detection module is used for online scene anomaly perception detection using the cloud ruler. 6. A scene abnormality perception index calculation device based on a deep hybrid cloud model, characterized in that it comprises a processor, a memory, and a computer program stored in the memory and running on the processor, and the processor executes The above computer program implements the steps of the method for calculating the scene abnormality perception index based on the deep hybrid cloud model according to any one of claims 1-4. 7. A computer-readable storage medium, characterized in that, the storage medium stores at least one segment of program, and the at least one segment of program is executed by the above-mentioned processor to realize any one of claims 1-4. The steps of the calculation method of scene anomaly perception index based on deep hybrid cloud model.

Images