CN117640434A

CN117640434A - Block chain node state detection method and device

Info

Publication number: CN117640434A
Application number: CN202311596072.2A
Authority: CN
Inventors: 李曼潇
Original assignee: Industrial and Commercial Bank of China Ltd ICBC
Current assignee: Industrial and Commercial Bank of China Ltd ICBC
Priority date: 2023-11-27
Filing date: 2023-11-27
Publication date: 2024-03-01

Abstract

The invention provides a block chain node state detection method and device, relates to the technical field of artificial intelligence, and can be applied to the technical field of finance or other technical fields. The block chain link point state detection method comprises the following steps: inputting the block chain link point characteristic information into a transaction amount prediction model created based on the historical node characteristic information and the corresponding actual transaction amount to obtain a predicted transaction amount; inputting the block chain link point characteristic information into an abnormality prediction model created based on the historical node characteristic information and the corresponding actual abnormality type to obtain a node abnormality type; and inputting the predicted transaction quantity, the node abnormality type and the block link point characteristic information into a state monitoring image model created based on the historical node characteristic information and the corresponding historical state monitoring image to obtain a node state monitoring image. The invention can improve the flexibility and the intelligent degree of the point state detection of the block chain links and fully utilize the network computing resources.

Description

Block chain node state detection method and device

Technical Field

The invention relates to the technical field of block chains, in particular to a method and a device for detecting the point state of a block chain link.

Background

The health degree inspection and fault monitoring of the blockchain nodes in the industry have cases of using a multi-mode learning model, but a traditional LSTM model is generally used, training source data are log files of applications, model output is a binary result of normal or abnormal nodes, and the defects of single data source and single training target are overcome in design.

Disclosure of Invention

The embodiment of the invention mainly aims to provide a block chain node state detection method and device, so as to improve the flexibility and the intelligent degree of block chain node state detection and fully utilize network computing resources.

In order to achieve the above object, an embodiment of the present invention provides a blockchain node status detection method, including:

inputting the block chain link point characteristic information into a transaction amount prediction model created based on the historical node characteristic information and the corresponding actual transaction amount to obtain a predicted transaction amount;

inputting the block chain link point characteristic information into an abnormality prediction model created based on the historical node characteristic information and the corresponding actual abnormality type to obtain a node abnormality type;

and inputting the predicted transaction quantity, the node abnormality type and the block link point characteristic information into a state monitoring image model created based on the historical node characteristic information and the corresponding historical state monitoring image to obtain a node state monitoring image.

In one embodiment, the condition monitoring image model includes an image text model and a decoding model;

inputting the predicted transaction amount, the node abnormality type and the block link point characteristic information into a state monitoring image model created based on historical node characteristic information and corresponding historical state monitoring images, wherein the obtaining the node state monitoring image comprises the following steps:

performing text feature fusion on the predicted transaction amount, the node anomaly type and the blockchain node feature information to obtain blockchain node fusion data;

inputting the blockchain node fusion data into an image text model to obtain node image characteristics;

and inputting the node image characteristics into a decoding model to obtain a node state monitoring image.

In one embodiment, the step of creating a state monitoring image model includes:

inputting the historical state monitoring image into a coding model to obtain historical node image characteristics;

training the coding model and the decoding model according to the historical node image characteristics and the historical state monitoring image to obtain coding model parameters and decoding model parameters respectively;

performing text feature fusion on the historical node feature information, the historical predicted transaction amount and the historical node anomaly type to obtain historical blockchain node fusion data;

Training the image text model according to the historical blockchain node fusion data and the historical node image characteristics to obtain image text model parameters;

iteratively adjusting the coding model parameters and the decoding model parameters according to the image text model parameters until a preset iteration condition is met;

and creating the decoding model according to the decoding model parameters, and creating the state monitoring image model according to the image text model parameters.

In one embodiment, the method further comprises:

and carrying out text feature fusion on the node log data and the node related data to obtain the block link point feature information.

In one embodiment, the step of creating a transaction amount prediction model includes:

inputting the historical node characteristic information into a transaction amount prediction model to obtain transaction amount prediction data;

and iteratively adjusting model parameters of the transaction amount prediction model according to the transaction amount prediction data and the corresponding actual transaction amount, and creating the transaction amount prediction model according to the model parameters of the transaction amount prediction model.

In one embodiment, the step of creating an anomaly prediction model includes:

inputting the historical node characteristic information into an anomaly prediction model to obtain anomaly prediction data;

Determining an abnormal loss function according to the abnormal prediction data and the corresponding actual abnormal type;

and iteratively adjusting model parameters of the abnormal prediction model according to the abnormal loss function, and creating the abnormal prediction model according to the model parameters of the abnormal prediction model.

The embodiment of the invention also provides a device for detecting the state of the block chain node, which comprises the following steps:

the transaction amount prediction module is used for inputting the block link point characteristic information into a transaction amount prediction model created based on the historical node characteristic information and the corresponding actual transaction amount to obtain a predicted transaction amount;

the node abnormality module is used for inputting the block link point characteristic information into an abnormality prediction model created based on the historical node characteristic information and the corresponding actual abnormality type to obtain a node abnormality type;

and the node state monitoring image module is used for inputting the predicted transaction quantity, the node abnormality type and the block link point characteristic information into a state monitoring image model created based on the historical node characteristic information and the corresponding historical state monitoring image to obtain a node state monitoring image.

The node state monitoring image module comprises:

the block chain node fusion data unit is used for carrying out text feature fusion on the predicted transaction quantity, the node abnormality type and the block chain node feature information to obtain block chain node fusion data;

the node image feature unit is used for inputting the blockchain node fusion data into an image text model to obtain node image features;

and the node state monitoring image unit is used for inputting the node image characteristics into the decoding model to obtain a node state monitoring image.

In one embodiment, the method further comprises: a state monitoring image model creation module;

the state monitoring image model creation module includes:

the historical node image characteristic unit is used for inputting the historical state monitoring image into the coding model to obtain historical node image characteristics;

the coding and decoding model parameter unit is used for training the coding model and the decoding model according to the historical node image characteristics and the historical state monitoring image to obtain coding model parameters and decoding model parameters respectively;

the historical blockchain node fusion data unit is used for carrying out text feature fusion on the historical node feature information, the historical predicted transaction amount and the historical node anomaly type to obtain historical blockchain node fusion data;

The image text model parameter unit is used for training the image text model according to the historical blockchain node fusion data and the historical node image characteristics to obtain image text model parameters;

the iteration unit is used for iteratively adjusting the coding model parameters and the decoding model parameters according to the image text model parameters until a preset iteration condition is met;

and the state monitoring image model creation unit is used for creating the decoding model according to the decoding model parameters and creating the state monitoring image model according to the image text model parameters.

In one embodiment, the method further comprises:

and the text feature fusion module is used for carrying out text feature fusion on the node log data and the node related data to obtain the block link point feature information.

In one embodiment, the method further comprises: a transaction amount prediction model creation module;

the transaction amount prediction model creation module includes:

the transaction amount prediction data unit is used for inputting the historical node characteristic information into a transaction amount prediction model to obtain transaction amount prediction data;

and the transaction amount prediction model creation unit is used for iteratively adjusting model parameters of the transaction amount prediction model according to the transaction amount prediction data and the corresponding actual transaction amount, and creating the transaction amount prediction model according to the model parameters of the transaction amount prediction model.

In one embodiment, the method further comprises: an anomaly prediction model creation module;

the abnormality prediction model creation module includes:

the abnormal prediction data unit is used for inputting the historical node characteristic information into an abnormal prediction model to obtain abnormal prediction data;

an abnormal loss function unit, configured to determine an abnormal loss function according to the abnormal prediction data and the corresponding actual abnormal type;

and the abnormal prediction model creation unit is used for iteratively adjusting the model parameters of the abnormal prediction model according to the abnormal loss function and creating the abnormal prediction model according to the model parameters of the abnormal prediction model.

The embodiment of the invention also provides electronic equipment, which comprises a memory, a processor and a computer program stored on the memory and running on the processor, wherein the processor realizes the steps of the block chain link point state detection method when executing the computer program.

The embodiment of the invention also provides a computer readable storage medium, on which a computer program is stored, which when being executed by a processor, realizes the steps of the block link point state detection method.

The embodiment of the invention also provides a computer program product, which comprises a computer program/instruction, wherein the computer program/instruction realizes the steps of the block link point state detection method when being executed by a processor.

According to the blockchain node state detection method and device, the blockchain node characteristic information is input into the transaction amount prediction model to obtain the predicted transaction amount, the blockchain node characteristic information is input into the abnormal prediction model to obtain the node abnormal type, and finally the predicted transaction amount, the node abnormal type and the blockchain node characteristic information are input into the state monitoring image model to obtain the node state monitoring image, so that the flexibility and the intelligent degree of blockchain node state detection can be improved, and network computing resources are fully utilized.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are needed in the description of the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a schematic diagram illustrating a block chain node status detection stage in an embodiment of the present invention;

FIG. 2 is a flow chart of a block chain node status detection method in an embodiment of the invention;

FIG. 3 is a flowchart of a specific application of a block chain node status detection method in an embodiment of the present invention;

FIG. 4 is a flow chart of creating a traffic prediction model in an embodiment of the invention;

FIG. 5 is a flow chart of creating an anomaly prediction model in an embodiment of the present invention;

FIG. 6 is a flow chart of creating a state monitoring image model in an embodiment of the invention;

FIG. 7 is a flowchart of S103 in an embodiment of the present invention;

FIG. 8 is a flow chart of creating and applying a transaction amount prediction model in another embodiment of the invention;

FIG. 9 is a flow chart for creating and applying an anomaly prediction model in an embodiment of the present invention;

FIG. 10 is a flow chart for creating a condition monitoring image model in accordance with another embodiment of the present invention;

FIG. 11 is a flowchart of obtaining a node status monitoring image in another embodiment of the present invention;

FIG. 12 is a block diagram of a block chain node status detection device in accordance with an embodiment of the present invention;

FIG. 13 is a block diagram of a block chain node status detection device in accordance with another embodiment of the present invention;

fig. 14 is a schematic block diagram of a system configuration of an electronic device 9600 of an embodiment of the present application.

Detailed Description

The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

Those skilled in the art will appreciate that embodiments of the invention may be implemented as a system, apparatus, device, method, or computer program product. Accordingly, the present disclosure may be embodied in the following forms, namely: complete hardware, complete software (including firmware, resident software, micro-code, etc.), or a combination of hardware and software.

The technical scheme of the invention obtains, stores, uses, processes and the like the data, which all meet the relevant regulations of national laws and regulations. The user information in the embodiment of the application is obtained through legal compliance approaches, and the user information is obtained, stored, used, processed and the like through client authorization and consent.

In view of the drawbacks of single data source and single training target in the prior art, the embodiment of the invention provides a blockchain node state detection method and device, which combines blockchain and multi-mode learning to realize an intelligent monitoring scheme of blockchain nodes, analyze source data including multidimensional indexes such as application log files, node memories, disks, networks, transaction overtime, node state monitoring images and the like, analyze association characteristics among various source data and align the characteristics of various source data in the multi-mode learning of blockchain point state monitoring. Besides predicting node abnormality, the downstream task of the multi-mode model can also predict the change of the transaction amount of the blockchain network, position whether the abnormal type of the blockchain link point is insufficient or unsatisfied with the resource consistency, generate a visualized node monitoring image and the like, and realize the blockchain intelligent operation and maintenance and risk prediction. The present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a schematic diagram illustrating a block chain node status detection stage according to an embodiment of the present invention. As shown in FIG. 1, the invention uses a multi-modal learning model, wherein training data sources comprise multi-dimensional text data such as node memory, disk, network, transaction timeout and the like, and block link point state monitoring graphics are used as training image data. The downstream task design is rich, and comprises the steps of predicting node resource abnormality or unsatisfied consistency, predicting blockchain network transaction amount change, generating a visualized node monitoring image and the like, so that the defects of single training data source and single training target of the traditional blockchain link point state monitoring are overcome, the visualized node monitoring image is generated more intelligently, and the cost of manpower development required by a traditional monitoring visual interface is reduced.

The block link point state monitoring of the invention is divided into a block link point training stage and a block link point state monitoring stage.

In the 'block chain link point training stage', the invention adjusts the parameters of the multi-mode model, including 5 intelligent contracts, respectively executes the characteristics of text characteristic extraction, image characteristic extraction and generation, image-text characteristic matching, learning block chain network transaction amount, learning block chain node insufficient resources or abnormal types of characteristics such as unsatisfied consistency, prepares training data for each intelligent contract, and continuously adjusts the parameters of the multi-mode model in multiple iterations, so that the objective function value of training is continuously optimized.

In the 'block chain link point state monitoring stage', the invention opens a plurality of downstream tasks, including 5 intelligent contracts, respectively executes text feature extraction, generates visualized node monitoring images, confirms image features matched with the text features, predicts block chain network transaction amount, predicts abnormal types such as insufficient or unsatisfied consistency of block chain node resources and the like.

FIG. 2 is a flow chart of a block chain node status detection method in an embodiment of the invention. Fig. 3 is a flowchart of a specific application of the block chain node status detection method in the embodiment of the present invention. As shown in fig. 2-3, the overall inputs of the monitoring system at the application stage include: the method comprises the steps of storing characteristic data such as a block chain node memory, a disk, a network, a transaction timeout and the like, logging of the block chain node, and finally outputting monitoring image data (actually multi-mode) comprising a plurality of indexes, wherein intermediate results comprise transaction quantity predicted values in a period of time, whether block chain node resources are insufficient and consistency is not met. The model training phase includes, but is not limited to, the above inputs. The block chain link point state detection method comprises the following steps:

s101: and inputting the block link point characteristic information into a transaction amount prediction model created based on the historical node characteristic information and the corresponding actual transaction amount to obtain a predicted transaction amount.

In one embodiment, before executing S101, the method further includes:

FIG. 4 is a flow chart of creating a traffic prediction model in an embodiment of the invention. FIG. 8 is a flow chart of creating and applying a transaction amount prediction model in another embodiment of the invention. As shown in fig. 4 and 8, the step of creating a transaction amount prediction model includes:

s201: and inputting the historical node characteristic information into a transaction amount prediction model to obtain transaction amount prediction data.

In specific implementation, the intelligent contract 1 executes the BERT model, extracts semantic features of the node logs, obtains some characteristics of the blockchain transaction, and inputs the characteristics into the next step after being spliced with other characteristics of the blockchain link point memory, the disk, the network, the transaction timeout and the like.

The intelligent contract 4 executes multi-layer multi-head transformer encoder calculation, in order to predict the transaction amount of the blockchain network at the k moment, the semantic features, the memory, the disk, the network, the transaction timeout and other monitoring data of the blockchain node log at the 0-k moment are input, and the model learns the distribution features of the monitoring data under the cross-latitude and cross-time conditions. The output encoding result is the input of all decoder second layer patents in the next step.

S202: and iteratively adjusting model parameters of the transaction amount prediction model according to the transaction amount prediction data and the corresponding actual transaction amount, and creating the transaction amount prediction model according to the model parameters of the transaction amount prediction model.

In particular, the smart contract 4 performs autoregressive multi-layer multi-headed transformer decoder (transducer decoder) calculations, and the model learns the correlation characteristics between the network transaction amount and the monitored data, and the correlation characteristics between the network transaction amounts at the front and rear moments. In the training stage, in order to predict that the transaction amount of the blockchain network at the moment k is close to the true value, a first layer of mask layer of a first decoder inputs a matrix after the block chain network transaction amount sequence at the moment 0 to the moment k-1 is subjected to position coding, and then the first layer of mask layer of each decoder inputs the output of the last decoder, so that after the normalization result finally output by the softmax after the multi-layer decoders, the normalization value corresponding to the predicted value which is closer to the true value is larger.

In order to predict the transaction amount of the blockchain network at the time of k, when executing S101, the first layer of mask schedule of the first decoder inputs a matrix after position encoding of the result with higher normalized value in the final output sequence of the first decoder at the time of k-1, and the post-decoder softmax outputs the distribution situation of the transaction amount predicted at the time of k, and selects the blockchain network transaction amount with higher normalized value as the predicted result (predicted transaction amount).

S102: and inputting the block chain link point characteristic information into an abnormality prediction model created based on the historical node characteristic information and the corresponding actual abnormality type to obtain the node abnormality type.

FIG. 5 is a flow chart of creating an anomaly prediction model in an embodiment of the present invention. FIG. 9 is a flow chart of creating and applying an anomaly prediction model in an embodiment of the present invention. As shown in fig. 5 and 9, the step of creating an anomaly prediction model includes:

s301: and inputting the historical node characteristic information into an anomaly prediction model to obtain anomaly prediction data.

In specific implementation, the intelligent contract 1 executes the BERT model, extracts semantic features of the node logs, and obtains some characteristics of the blockchain transaction. And (3) splicing the semantic features, memory, disk, network, transaction timeout and other monitoring features of the block link point log together and then inputting the same into the next step.

S302: and determining an abnormal loss function according to the abnormal prediction data and the corresponding actual abnormal type.

In specific implementation, the intelligent contract 5 executes multi-layer multi-head transformer encoder calculation, and the model learns the associated distribution characteristics of abnormal types such as insufficient or unsatisfied consistency of the monitoring data and the node resources. In the training stage, according to the semantic features of the link point log of the input block, the memory, the disk, the network, the transaction timeout and the like, the parameters of the multi-head multilayer transformer encoder are adjusted, and the target loss function is defined as follows:

Wherein x is _i Is a high-dimensional sample data (a multi-mode data format uniquely designed by the invention) composed of log semantic features, memory, disk, network, transaction timeout and other dimensional data obtained in the ith block link point monitoring, wherein d (x _i ,x _j ) Is x _i 、x _j The distance between two samples in the high-dimensional feature space, S _normal Is a sample set of normal nodes and,is a set of nodes of various different exception types, size (S) is the size of the sample set. The training objective is to adjust model parameters and change the distribution of samples in the feature space, so that the objective loss function value is as small as possible, even if the distance of the output node feature codes meets the following conditions: the feature coding distances between the normal node samples are close, the feature coding distances between the abnormal node samples of the same type are close, the feature coding distances between the abnormal node samples of different types are far, and the feature coding distances between the normal node samples and the abnormal node samples are far.

S303: and iteratively adjusting model parameters of the abnormal prediction model according to the abnormal loss function, and creating the abnormal prediction model according to the model parameters of the abnormal prediction model.

When executing S102, determining whether the sample node to be predicted is abnormal or inconsistent with other blockchain nodes in the network by comparing the distances between the sample feature codes to be predicted and feature codes of normal node samples and abnormal node samples of various types, and further obtaining the node abnormal type.

S103: and inputting the predicted transaction quantity, the node abnormality type and the block link point characteristic information into a state monitoring image model created based on the historical node characteristic information and the corresponding historical state monitoring image to obtain a node state monitoring image.

Wherein the state monitoring image model comprises an image text model and a decoding model.

FIG. 6 is a flow chart of creating a state monitoring image model in an embodiment of the invention. FIG. 10 is a flow chart for creating a condition monitoring image model in another embodiment of the invention. As shown in fig. 6 and 10, the step of creating the state monitoring image model includes:

s401: and inputting the historical state monitoring image into a coding model to obtain the image characteristics of the historical nodes.

In specific implementation, the coding parameters and decoding parameters of the discrete variation self-encoder dVAE model are trained, the historical state monitoring image is compressed, and meanwhile, the visual characteristics of the high-frequency target image, namely a discretized image token sequence (historical node image characteristics) are extracted and used as the input characteristics of the original image.

The intelligent contract 2 performs dVAE model coding calculation to code the input monitoring image data x into a probability distribution q on the image feature space _φ (z|x) instead of a single point z, then sampling an hidden variable z-q according to the distribution _φ (z|x) an image token z (history node image feature) is obtained.

S402: and training the coding model and the decoding model according to the historical node image characteristics and the historical state monitoring image to obtain coding model parameters and decoding model parameters respectively.

In specific implementation, the relation between the image token and the image block is recorded, and dVAE model decoding calculation is executed. Recovering high-resolution picture through image token, and continuously adjusting parameter q of coding model _φ (z|x), parameter p of decoding model _θ (x|z) such that the dVAE model decoded output image x' approaches the expected image (historical state monitoring image) to obtain the first stage pre-trained model parameters, temporarily fixing this part of the model parameters q _φ (y,z|x)＝q _φ (z|x)、p _θ (x|y,z)＝p _θ (x|z) and doing the next training.

S403: and carrying out text feature fusion on the historical node feature information, the historical predicted transaction amount and the historical node abnormal type to obtain historical block link point fusion data.

The history node characteristic information comprises node logs, node memories, disks, networks, transaction timeout and other characteristics

In specific implementation, the intelligent contract 1 executes BERT token operation to extract text features of the blockchain node log, performs text feature fusion calculation with other text features such as blockchain link point memory, disk, network, transaction timeout and the like, historical predicted transaction amount and historical node anomaly type, and transmits the fused text features y (historical blockchain link point fusion data) to an autoregressive transformation model of the intelligent contract 3 Training of the autoregressive transducer model was performed.

S404: and training the image text model according to the historical blockchain node fusion data and the historical node image characteristics to obtain image text model parameters.

In particular, smart contract 2 adjusts an autoregressive transducer modelIs to maximize the likelihood function of the input text feature data y with the desired output image feature data x, the objective function using a near-netA similar lower variation bound is defined as follows:

wherein the autoregressive transducer model will be as followsThe standard normal distribution relation associates the text coding data y with the picture coding data z (historical node image characteristics), the model output generates a plurality of picture coding data z at one time, the picture coding data z is ordered in a rerank mode, finally, the image token with the highest matching score with the text token is selected as a training result, and model parameters are continuously adjusted through training, so that the model obtains a proper picture-text matching relation with high probability, and the model parameters of the second stage are obtained. Temporary fixation->And (5) training the image text model parameters in the next stage.

S405: and iteratively adjusting the coding model parameters and the decoding model parameters according to the image text model parameters until a preset iteration condition is met.

In practice, the intelligent contract 2 adjusts parameters of the dVAE encoding and dVAE decoding models, and the training objective function uses an approximate variant lower bound, defined as follows:

wherein, dVAE codes the input monitoring image x according to q _φ The (y, z|x) Gaussian mixture model is used for calculating distribution, so that the original data x (historical state monitoring image) of an input image can be converted into output according to any function distribution relation, and image coding data z and text coding data y are obtained. dVAE decoding according to p _θ The (x|y, z) functional relationship converts the text encoding data y and the image encoding data z into an output image x ', x' distribution characteristics approximate to the original outputInto the distribution characteristics of the image x. The training process continuously adjusts the image encoding space so that the actual output is continuously close to the expected output. Introduction of KL divergence by objective functionSo as to ensure continuity and integrity of image feature token space, and the regular term proportion parameter beta of KL divergence is continuously learned in model training.

S406: and creating the decoding model according to the decoding model parameters, and creating the state monitoring image model according to the image text model parameters.

In the specific implementation, the parameter adjustment of the model is completed in multiple rounds by iteration, when KL distance items of dVAE coding output distribution and autoregressive transform model output distribution are as small as possible, posterior probability items of dVAE decoding output are as large as possible, the lower limit of likelihood probability of an objective function can be improved as much as possible, and the effect of the model is better. After the iteration is completed, the decoding model and the state monitoring image model are respectively created by adopting model parameters in the last iteration.

Fig. 7 is a flowchart of S103 in the embodiment of the present invention. FIG. 11 is a flow chart of obtaining a node status monitoring image in another embodiment of the invention. As shown in fig. 7 and 11, S103 includes:

s501: and carrying out text feature fusion on the predicted transaction quantity, the node anomaly type and the blockchain node feature information to obtain blockchain link point fusion data.

In specific implementation, the intelligent contract 1 executes the BERT model to extract text features of the blockchain node logs, performs text feature fusion calculation with other text features such as blockchain link point memories, magnetic disks, networks, transaction overtime and the like, historical predicted transaction amounts and historical node anomaly types, and transmits the fused text features to the intelligent contract 3.

S502: and inputting the blockchain node fusion data into an image text model to obtain node image characteristics.

In specific implementation, the intelligent contract 3 performs image-text feature matching, and selects an image feature token (node image feature) closest to the text feature token.

S503: and inputting the node image characteristics into a decoding model to obtain a node state monitoring image.

In specific implementation, the intelligent contract 2 obtains an image patch (image block) according to the selected image feature token, and inputs a series of image patches to execute dVAE decoding calculation to generate a monitoring image (node image feature).

The blockchain node state detection method shown in fig. 2 may be implemented by a computer. As can be seen from the flow chart shown in fig. 2, the blockchain node state detection method in the embodiment of the invention inputs the blockchain node point characteristic information into the transaction amount prediction model to obtain the predicted transaction amount, inputs the blockchain node point characteristic information into the anomaly prediction model to obtain the node anomaly type, and finally inputs the predicted transaction amount, the node anomaly type and the blockchain node characteristic information into the state monitoring image model to obtain the node state monitoring image, so that the flexibility and the intelligent degree of blockchain node point state detection can be improved, and network computing resources can be fully utilized.

In summary, the block link point state detection method provided by the embodiment of the invention has the following beneficial effects:

(1) The invention adopts a multi-mode large model based on a transducer, uses a BERT token as semantic analysis and extraction, uses dVAE as image compression coding and image generation, uses an autoregressive transducer as image-text matching task, solves the problems that the traditional method uses LSTM model deep stacking, introduces the problems of slow coding calculation, low training efficiency, model gradient problem, incapability of further improving model effect of a simple multi-layer stacking basic model and the like caused by a serial structure, and the dVAE uses a Gaussian mixture model to support the matching of various complex forms of an original image and an image characteristic token, so that the method has more flexibility;

(2) The model adopted by the invention is richer in information used in feature extraction and richer in downstream tasks, not only can realize the functions of predicting whether the abnormal type of the node is insufficient or unsatisfied with consistency, predicting the change of the transaction amount of the block chain network, comparing the consistency of the block chain node account book and the like, but also can obtain the visualized node monitoring image corresponding to the monitoring text information through the trained graph-text matching model and the image generating model of the multi-mode large model, thereby saving the workload of manually developing the block chain link point state monitoring page and further improving the intelligent degree of the block chain link point state monitoring;

(3) The invention uses the block chain intelligent contract and training model data consensus block-falling mechanism to complete the training design of the multi-round model, so that the adjustment of the multi-mode large model is flexible, each training stage has no binding relation with a specific node server, and the computing resources in the block chain network are fully utilized in the model training and application stages.

Based on the same inventive concept, the embodiment of the invention also provides a block chain node state detection device, and because the principle of solving the problem of the device is similar to that of the block chain node state detection method, the implementation of the device can refer to the implementation of the method, and the repetition is omitted.

Fig. 12 is a block diagram of a block chain node status detection device in an embodiment of the present invention. As shown in fig. 12, the blockchain node status detection device includes:

the node state monitoring image module comprises:

the state monitoring image model creation module includes:

In one embodiment, the method further comprises:

the transaction amount prediction model creation module includes:

the abnormality prediction model creation module includes:

Fig. 13 is a block diagram of a block chain node status detection device in accordance with another embodiment of the present invention. As shown in fig. 13, in practical application, the block link point state detection device includes individual intelligent contracts and basic components.

The intelligent contract 1 is constructed on a BERT model foundation assembly and comprises a node state monitoring image module, a state monitoring image model creation module and a text feature fusion module, supports text semantic analysis and extraction, outputs a series of text token, and can be used for analyzing log files of blockchain nodes. The BERT model is the basis for the task of visual block link point state monitoring image generation, and other series of related block chain node log file analysis.

The intelligent contract 2 is constructed on a dVAE model basic component and comprises a node state monitoring image module and a state monitoring image model creation module, when in coding, the model divides a complete image into a series of image patches, extraction of local features and global features of the image is supported, an image feature token is obtained after compression, and when in decoding, the model recovers the complete image from the image feature token. The dVAE model is the basis for the generation of a visual block link point state monitoring image.

The intelligent contract 3 is constructed on an autoregressive transformation model base component and comprises a node state monitoring image module and a state monitoring image model creation module, supports image-text matching, and provides a text token model capable of providing the closest image token. The autoregressive transducer model supporting graphic matching is the basis for the generation of the state monitoring image of the visual block chain link.

The intelligent contract 4 is constructed on an autoregressive transducer model base component and comprises a transaction amount prediction module and a transaction amount prediction model creation module, and is used for supporting blockchain network transaction amount prediction.

The intelligent contract 5 is constructed on a transformer encoder model foundation component and comprises a node exception module and an exception prediction model creation module, and the prediction of insufficient or unsatisfied consistency judgment of block chain node resources is supported.

In summary, the blockchain node state detection device of the embodiment of the invention inputs the blockchain link point characteristic information into the transaction amount prediction model to obtain the predicted transaction amount, inputs the blockchain link point characteristic information into the anomaly prediction model to obtain the node anomaly type, and finally inputs the predicted transaction amount, the node anomaly type and the blockchain node characteristic information into the state monitoring image model to obtain the node state monitoring image, so that the flexibility and the intelligent degree of blockchain link point state detection can be improved, and network computing resources are fully utilized.

Fig. 14 is a schematic block diagram of a system configuration of an electronic device 9600 of an embodiment of the present application. As shown in fig. 14, the electronic device 9600 may include a central processor 9100 and a memory 9140; the memory 9140 is coupled to the central processor 9100. Notably, this fig. 14 is exemplary; other types of structures may also be used in addition to or in place of the structures to implement telecommunications functions or other functions.

In one embodiment, the block link point state detection method functionality may be integrated into the CPU 9100. The central processor 9100 may be configured to perform the following control:

As can be seen from the above description, the blockchain node state detection method provided by the application inputs the blockchain node point characteristic information into the transaction amount prediction model to obtain the predicted transaction amount, inputs the blockchain node point characteristic information into the anomaly prediction model to obtain the node anomaly type, and finally inputs the predicted transaction amount, the node anomaly type and the blockchain node characteristic information into the state monitoring image model to obtain the node state monitoring image, so that the flexibility and the intelligent degree of blockchain node state detection can be improved, and network computing resources are fully utilized.

In another embodiment, the block link point state detection device may be configured separately from the cpu 9100, for example, the block link point state detection device may be configured as a chip connected to the cpu 9100, and the function of the block link point state detection method is implemented by the control of the cpu.

As shown in fig. 14, the electronic device 9600 may further include: a communication module 9110, an input unit 9120, an audio processor 9130, a display 9160, and a power supply 9170. It is noted that the electronic device 9600 need not include all of the components shown in fig. 14; in addition, the electronic device 9600 may further include components not shown in fig. 14, and reference may be made to the related art.

As shown in fig. 14, the central processor 9100, sometimes referred to as a controller or operational control, may include a microprocessor or other processor device and/or logic device, which central processor 9100 receives inputs and controls the operation of the various components of the electronic device 9600.

The memory 9140 may be, for example, one or more of a buffer, a flash memory, a hard drive, a removable media, a volatile memory, a non-volatile memory, or other suitable device. The information about failure may be stored, and a program for executing the information may be stored. And the central processor 9100 can execute the program stored in the memory 9140 to realize information storage or processing, and the like.

The input unit 9120 provides input to the central processor 9100. The input unit 9120 is, for example, a key or a touch input device. The power supply 9170 is used to provide power to the electronic device 9600. The display 9160 is used for displaying display objects such as images and characters. The display may be, for example, but not limited to, an LCD display.

The memory 9140 may be a solid state memory such as Read Only Memory (ROM), random Access Memory (RAM), SIM card, etc. But also a memory which holds information even when powered down, can be selectively erased and provided with further data, an example of which is sometimes referred to as EPROM or the like. The memory 9140 may also be some other type of device. The memory 9140 includes a buffer 9141 (sometimes referred to as a buffer memory). The memory 9140 may include an application/function storage portion 9142, the application/function storage portion 9142 storing application programs and function programs or a flow for executing operations of the electronic device 9600 by the central processor 9100.

The memory 9140 may also include a data store 9143, the data store 9143 for storing data, such as contacts, digital data, pictures, sounds, and/or any other data used by an electronic device. The driver storage portion 9144 of the memory 9140 may include various drivers of the electronic device for communication functions and/or for performing other functions of the electronic device (e.g., messaging applications, address book applications, etc.).

The communication module 9110 is a transmitter/receiver 9110 that transmits and receives signals via an antenna 9111. A communication module (transmitter/receiver) 9110 is coupled to the central processor 9100 to provide input signals and receive output signals, as in the case of conventional mobile communication terminals.

Based on different communication technologies, a plurality of communication modules 9110, such as a cellular network module, a bluetooth module, and/or a wireless local area network module, etc., may be provided in the same electronic device. The communication module (transmitter/receiver) 9110 is also coupled to a speaker 9131 and a microphone 9132 via an audio processor 9130 to provide audio output via the speaker 9131 and to receive audio input from the microphone 9132 to implement usual telecommunications functions. The audio processor 9130 can include any suitable buffers, decoders, amplifiers and so forth. In addition, the audio processor 9130 is also coupled to the central processor 9100 so that sound can be recorded locally through the microphone 9132 and sound stored locally can be played through the speaker 9131.

The embodiment of the present invention further provides a computer readable storage medium capable of implementing all the steps in the block link state detection method in which the execution subject in the above embodiment is a server or a client, where the computer readable storage medium stores a computer program, and when the computer program is executed by a processor, the computer program implements all the steps in the block link state detection method in the above embodiment, for example, the processor implements the following steps when executing the computer program:

In summary, the computer readable storage medium of the embodiment of the invention inputs the block link point characteristic information into the transaction amount prediction model to obtain the predicted transaction amount, inputs the block link point characteristic information into the anomaly prediction model to obtain the node anomaly type, and finally inputs the predicted transaction amount, the node anomaly type and the block link node characteristic information into the state monitoring image model to obtain the node state monitoring image, so that the flexibility and the intelligent degree of block link point state detection can be improved, and network computing resources are fully utilized.

The embodiment of the present invention further provides a computer program product capable of implementing all the steps in the block link point state detection method in which the execution subject is a server or a client in the above embodiment, where the computer program product includes a computer program/instruction, and the computer program/instruction implements all the steps in the block link point state detection method in the above embodiment when executed by a processor, for example, the processor implements the following steps when executing the computer program:

In summary, the computer program product of the embodiment of the invention inputs the characteristic information of the block chain link points into the transaction quantity prediction model to obtain the predicted transaction quantity, inputs the characteristic information of the block chain link points into the abnormality prediction model to obtain the node abnormality type, and finally inputs the predicted transaction quantity, the node abnormality type and the characteristic information of the block chain nodes into the state monitoring image model to obtain the node state monitoring image, so that the flexibility and the intelligent degree of the detection of the block chain link points can be improved, and the network computing resources are fully utilized.

In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments. In particular, for a hardware+program class embodiment, the description is relatively simple, as it is substantially similar to the method embodiment, as relevant see the partial description of the method embodiment.

The foregoing describes specific embodiments of the present disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims can be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Although the present application provides method operational steps as described in the examples or flowcharts, more or fewer operational steps may be included based on conventional or non-inventive labor. The order of steps recited in the embodiments is merely one way of performing the order of steps and does not represent a unique order of execution. When implemented by an actual device or client product, the instructions may be executed sequentially or in parallel (e.g., in a parallel processor or multi-threaded processing environment) as shown in the embodiments or figures.

Although the present description provides method operational steps as described in the examples or flowcharts, more or fewer operational steps may be included based on conventional or non-inventive means. The order of steps recited in the embodiments is merely one way of performing the order of steps and does not represent a unique order of execution. When implemented in an actual device or end product, the instructions may be executed sequentially or in parallel (e.g., in a parallel processor or multi-threaded processing environment, or even in a distributed data processing environment) as illustrated by the embodiments or by the figures. The terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, it is not excluded that additional identical or equivalent elements may be present in a process, method, article, or apparatus that comprises a described element.

For convenience of description, the above devices are described as being functionally divided into various modules, respectively. Of course, when implementing the embodiments of the present disclosure, the functions of each module may be implemented in the same or multiple pieces of software and/or hardware, or a module that implements the same function may be implemented by multiple sub-modules or a combination of sub-units, or the like. The above-described apparatus embodiments are merely illustrative, for example, the division of the units is merely a logical function division, and there may be additional divisions when actually implemented, for example, multiple units or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or units, which may be in electrical, mechanical or other form.

Those skilled in the art will also appreciate that, in addition to implementing the controller in a pure computer readable program code, it is well possible to implement the same functionality by logically programming the method steps such that the controller is in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers, etc. Such a controller can be regarded as a hardware component, and means for implementing various functions included therein can also be regarded as a structure within the hardware component. Or even means for achieving the various functions may be regarded as either software modules implementing the methods or structures within hardware components.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

In one typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.

The memory may include volatile memory in a computer-readable medium, random Access Memory (RAM) and/or nonvolatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of computer-readable media.

Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transmission media), such as modulated data signals and carrier waves.

It will be appreciated by those skilled in the art that embodiments of the present description may be provided as a method, system, or computer program product. Accordingly, the present specification embodiments may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present description embodiments may take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

Embodiments in this specification may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The embodiments of the specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.

In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments. In particular, for system embodiments, since they are substantially similar to method embodiments, the description is relatively simple, as relevant to see a section of the description of method embodiments. In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the embodiments of the present specification. In this specification, schematic representations of the above terms are not necessarily directed to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, the different embodiments or examples described in this specification and the features of the different embodiments or examples may be combined and combined by those skilled in the art without contradiction.

The foregoing is merely an example of an embodiment of the present disclosure and is not intended to limit the embodiment of the present disclosure. Various modifications and variations of the illustrative embodiments will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, or the like, which is within the spirit and principles of the embodiments of the present specification, should be included in the scope of the claims of the embodiments of the present specification.

Claims

1. A method for detecting the state of a block link, comprising:

2. The block link point state detection method of claim 1, wherein the state monitoring image model comprises an image text model and a decoding model;

3. The method of segment link point state detection according to claim 2, wherein the step of creating a state monitoring image model comprises:

4. The block link point state detection method of claim 1, further comprising:

5. The method of block link point state detection according to claim 1, wherein the step of creating a traffic prediction model comprises:

6. The block link point state detection method of claim 1, wherein the step of creating an anomaly prediction model comprises:

7. A block link point state detection device, comprising:

8. An electronic device comprising a memory, a processor, and a computer program stored on the memory and running on the processor, wherein the processor, when executing the computer program, implements the steps of the blockchain node state detection method of any of claims 1 to 6.

9. A computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the steps of the blockchain node state detection method of any of claims 1 to 6.

10. A computer program product comprising computer programs/instructions which when executed by a processor implement the steps of the blockchain node state detection method of any of claims 1 to 6.