CN114661010B

CN114661010B - Driving detection processing method based on artificial intelligence and cloud platform

Info

Publication number: CN114661010B
Application number: CN202210263019.XA
Authority: CN
Inventors: 兰立景; 李俊
Original assignee: Hangzhou Hesmore Information Technology Co ltd; Nanchang Hesi Information Technology Co ltd; Beijing Hesi Information Technology Co Ltd
Current assignee: Hangzhou Hesmore Information Technology Co ltd; Nanchang Hesi Information Technology Co ltd; Beijing Hesi Information Technology Co Ltd
Priority date: 2022-03-17
Filing date: 2022-03-17
Publication date: 2023-05-09
Anticipated expiration: 2042-03-17
Also published as: CN114661010A

Abstract

The invention discloses a drive detection processing method and a cloud platform based on artificial intelligence, and relates to the technical field of artificial intelligence, wherein the drive detection processing method based on the artificial intelligence comprises the following steps: responding to a system control instruction, and acquiring an unlabeled system operation feature vector matched with a target industrial system; forming an initial system operation characteristic vector set matched with a corresponding system operation environment; invoking a system control instruction generation model for countertraining by using a historical system running state training sample, and generating a control instruction for the initial system running feature vector set through the system control instruction generation model to obtain a control instruction of a target industrial system; by outputting the control instruction of the target industrial system, compared with the prior art that each industrial system is controlled manually, the control instruction of the system can be used for generating a model for accurate automatic control.

Description

Driving detection processing method based on artificial intelligence and cloud platform

Technical Field

The invention relates to the technical field of artificial intelligence, in particular to a drive detection processing method based on artificial intelligence and a cloud platform.

Background

Artificial intelligence (Artificial Intelligence), english is abbreviated AI. It is a new technical science for researching, developing theory, method, technology and application system for simulating, extending and expanding human intelligence. With the development of industrial production, more and more industrial systems are involved in artificial intelligence, and for these engineering-level KVM applications, the individual industrial systems (e.g., industrial computers) are typically operated by professional operators. However, due to the types and systems of the related industrial systems are more and more complicated, even if the running state of each flow of each industrial system is relatively fixed under the running scene of the industrial system, the running states of each industrial system are not completely consistent, and the start-stop operation of each industrial system is very complicated and easy to make mistakes by manually detecting the start-stop operation of each industrial system based on the HDMI interface drive of the main control system.

In view of this, it would be desirable to provide an artificial intelligence based drive detection processing scheme that can be automated.

Disclosure of Invention

The invention provides a cloud platform for driving detection processing method based on artificial intelligence.

In a first aspect, an embodiment of the present invention provides an artificial intelligence based drive detection processing method, applied to a computer system, where the computer system is communicatively connected to a plurality of industrial systems, the method includes:

Responding to a system control instruction, and acquiring an unlabeled system operation feature vector matched with a target industrial system;

determining different system operation environments corresponding to the label-free system operation feature vectors matched with the target industrial system respectively;

determining HDMI states matched with different system operating environments based on the different system operating environments;

based on the HDMI state, performing feature enhancement processing on the unlabeled system operation feature vectors matched with the target industrial system to form an initial system operation feature vector set matched with the corresponding system operation environment;

invoking a system control instruction generation model for countertraining by using a historical system running state training sample, and generating a control instruction for an initial system running feature vector set through the system control instruction generation model to obtain a control instruction of a target industrial system, wherein the system control instruction generation model comprises a feature extraction layer, a feature selection layer and an information filtering layer;

and outputting a control instruction of the target industrial system.

Optionally, the method further comprises:

acquiring characteristics of a target industrial system set and historical parameters of a system starting environment;

obtaining a system operation feature vector set matched with a system control instruction generation model according to the features of a target industrial system set and the historical parameters of a system starting environment, wherein the system operation feature vector set comprises sample features of different historical system operation states;

Determining the use environment of a system control instruction generation model, and determining sliding windows with different lengths according to the use environment of the system control instruction generation model;

determining time node data carried by different sample data in a system operation feature vector set;

according to a preset iterative algorithm matched with a system control instruction generation model and sliding windows with different lengths, processing time node data carried by different sample data in a system operation feature vector set, and determining sample system control data sets of different historical system operation states matched with the system control instruction generation model, wherein the sample system control data sets comprise at least one group of sample system control data;

training a system control instruction generation model according to a sample system control data set, and determining super parameters of a feature extraction layer in the system control instruction generation model;

training a system control instruction generation model according to a sample system control data set, and determining hyper-parameters of a feature selection layer in the system control instruction generation model;

training a system control instruction generation model according to the sample system control data set, and determining hyper-parameters of an information filtering layer in the system control instruction generation model so as to predict the instruction of a target industrial system in a system starting environment through the system control instruction generation model.

Optionally, training the system control instruction generation model according to the sample system control data set, and determining the hyper-parameters of the feature extraction layer in the system control instruction generation model, including:

generating a feature extraction layer in the model through a system control instruction, and processing a sample system control data set to determine initial parameters of the feature extraction layer;

responding to the initial parameters of the feature extraction layer, processing a sample system control data set through the feature extraction layer, and determining iteration parameters of the feature extraction layer;

determining a loss function corresponding to the feature extraction layer;

according to the iteration parameters of the feature extraction layer, carrying out iteration update on the parameters of the feature extraction layer;

until the loss function of the feature extraction layer reaches a corresponding preset convergence threshold, and based on parameters in the feature extraction layer, feature selection vectors of each sample in the sample system control data set can be extracted.

Optionally, training the system control instruction generation model according to the sample system control data set, and determining the hyper-parameters of the feature selection layer in the system control instruction generation model, including:

generating a feature selection layer in the model through a system control instruction, and processing a sample system control data set to determine initial parameters of the feature selection layer;

Substituting different sample system control data in the sample system control data set into a loss function corresponding to the feature selection layer;

determining corresponding iteration parameters of the feature selection layer when the loss function meets a corresponding preset convergence threshold;

determining a loss function corresponding to the feature selection layer;

according to the iteration parameters of the feature selection layer, carrying out iteration update on the parameters of the feature selection layer;

and determining instruction output results of different samples based on the corresponding sample labels and the feature selection vector of each sample until the loss function of the feature selection layer reaches a corresponding preset convergence threshold.

Optionally, training the system control instruction generation model according to the sample system control data set, and determining the super parameters of the information filtering layer in the system control instruction generation model, including:

responding to the initial parameters of the information filtering layer, processing the sample system control data set through the information filtering layer, and determining the iteration parameters of the information filtering layer;

according to the iteration parameters of the information filtering layer, the parameters of the information filtering layer are iteratively updated through the sample system control data set, so that the historical system running states corresponding to different samples in the sample set are judged.

Optionally, the method further comprises:

substituting different sample system control data in the sample system control data set into a loss function corresponding to the information filtering layer;

determining corresponding iteration parameters of the information filter layer when the loss function meets a corresponding preset convergence threshold;

determining a loss function corresponding to the information filtering layer;

according to the iteration parameters of the information filter layer, carrying out iteration update on the parameters of the information filter layer;

and judging the historical system running states corresponding to different samples in the sample set based on the parameters in the information filter layer until the loss function of the information filter layer reaches the corresponding preset convergence threshold.

Optionally, the computer system is further communicatively connected to both the authentication server and the user login system, and the method includes, prior to obtaining the unlabeled system operation feature vector matching the target industrial system in response to the system control instruction:

the method comprises the steps of obtaining user operation data uploaded by a verification server, and carrying out safety detection on system control instructions stored in a control instruction temporary storage space, wherein the control instruction temporary storage space comprises a target industrial system control instruction set, the target industrial system control instruction set comprises at least one system control instruction, and different system control instructions are respectively generated by different user login systems;

If a first system control instruction in at least one system control instruction passes the safety detection and the first system control instruction is the system control instruction with the most effective reference value in the target industrial system control instruction set, acquiring a hash parameter corresponding to the first system control instruction, acquiring an encrypted fingerprint carried by user operation data, and acquiring a first key corresponding to the verification server;

decrypting the encrypted fingerprint based on the first key to obtain first encrypted information corresponding to the encrypted fingerprint;

operating the user operation data based on the password hash function to obtain second encryption information corresponding to the user operation data;

if the first encryption information is the same as the second encryption information, the user operation data is verified and passed, and an operation data vector is generated based on the verified user operation data;

generating a vector label according to the hash parameter, generating a second system control instruction according to the vector label and the operation data vector, and adding the second system control instruction to a target industrial system control instruction set to obtain an updated control instruction temporary storage space;

traversing the second system control instruction in the user login system cluster to enable other user login systems except the user login system generating the second system control instruction in the user login system cluster, and respectively caching the second system control instruction into the belonging temporary storage space;

Acquiring the number of system control instructions contained in the updated control instruction temporary storage space, determining a user login system corresponding to each system control instruction in the updated control instruction temporary storage space, and acquiring credibility weights matched with the user login system;

updating the confidence threshold value corresponding to each system control instruction in the updated control instruction temporary storage space based on the number and the credibility weight of the system control instructions contained in the updated control instruction temporary storage space;

marking an instruction valid tag on the system control instruction with the updated confidence coefficient threshold value larger than the preset threshold value, and obtaining the current confidence coefficient corresponding to the system control instruction with the instruction valid tag;

if the difference between the current confidence coefficient and the confidence coefficient corresponding to the most effective reference value in the control instruction safety space is within a preset effective range, adding a system control instruction with an instruction effective label into the control instruction safety space;

if the difference between the current confidence coefficient and the confidence coefficient corresponding to the most effective reference value in the control instruction safety space is outside a preset effective range, performing confidence coefficient updating on the system control instruction with the instruction effective label, and adding the updated system control instruction with the instruction effective label into the control instruction safety space, wherein the control instruction safety space is used for storing all the system control instructions marked with the instruction effective label.

Optionally, the control instruction temporary storage space comprises a plurality of system control instruction sets, and the plurality of system control instruction sets comprise a target industrial system control instruction set;

the method for carrying out safety detection on the system control instruction stored in the control instruction temporary storage space comprises the following steps:

acquiring a plurality of system control instruction sets from a control instruction temporary storage space, and acquiring initial system control instruction numbers corresponding to the plurality of system control instruction sets respectively;

the method comprises the steps of sequencing a plurality of system control instruction sets based on the initial system control instruction number, and sequentially carrying out safety detection on the system control instructions contained in each system control instruction set based on the sequencing order of each system control instruction set.

Optionally, if the first system control instruction in the at least one system control instruction passes the security detection and the first system control instruction is a system control instruction with the most effective reference value in the target industrial system control instruction set, acquiring the hash parameter corresponding to the first system control instruction includes:

if a target industrial system control instruction set with the system control instructions passing through the safety detection exists in the plurality of system control instruction sets, taking the system control instruction with the most effective reference value in the target industrial system control instruction set as a first system control instruction, and acquiring a hash parameter corresponding to the first system control instruction.

In a second aspect, an embodiment of the present invention provides an artificial intelligence based drive detection processing cloud platform, applied to a computer system, where the computer system is communicatively connected to a plurality of industrial systems, the cloud platform includes:

the system comprises an acquisition module, a control module and a control module, wherein the acquisition module is used for responding to a system control instruction and acquiring a label-free system operation feature vector matched with a target industrial system;

the determining module is used for determining different system operation environments corresponding to the label-free system operation feature vectors matched with the target industrial system respectively; determining HDMI states matched with different system operating environments based on the different system operating environments; based on the HDMI state, performing feature enhancement processing on the unlabeled system operation feature vectors matched with the target industrial system to form an initial system operation feature vector set matched with the corresponding system operation environment;

the system control instruction generation model comprises a feature extraction layer, a feature selection layer and an information filtering layer; and outputting a control instruction of the target industrial system.

Compared with the prior art, the invention has the beneficial effects that: by adopting the driving detection processing method based on the artificial intelligence, which is provided by the embodiment of the invention, the operation feature vector of the label-free system matched with the target industrial system is obtained by responding to the system control instruction; determining different system operation environments corresponding to the operation feature vectors of the label-free system matched with the target industrial system respectively; further, based on different system operation environments, determining HDMI states matched with the different system operation environments; then, based on the HDMI state, carrying out feature enhancement processing on the unlabeled system operation feature vectors matched with the target industrial system to form an initial system operation feature vector set matched with the corresponding system operation environment; then, a system control instruction generation model for countertraining by using a historical system running state training sample is called, and control instruction generation is carried out on an initial system running feature vector set through the system control instruction generation model to obtain a control instruction of a target industrial system, wherein the system control instruction generation model comprises a feature extraction layer, a feature selection layer and an information filtering layer; finally, a control instruction of the target industrial system is output, and through the steps, the automatic control of the industrial system is realized by skillfully utilizing a system control instruction generation model.

Drawings

In order to more clearly illustrate the technical solution of the embodiments of the present invention, the drawings that are required to be used in the embodiments will be briefly described. It is appreciated that the following drawings depict only certain embodiments of the invention and are therefore not to be considered limiting of its scope. Other relevant drawings may be made by those of ordinary skill in the art without undue burden from these drawings.

FIG. 1 is an interactive schematic diagram of an artificial intelligence based drive detection processing system according to an embodiment of the present invention;

FIG. 2 is a schematic flow chart of a driving detection processing method based on artificial intelligence according to an embodiment of the present invention;

FIG. 3 is a schematic block diagram of a driving detection processing cloud platform based on artificial intelligence according to an embodiment of the present invention;

fig. 4 is a schematic block diagram of a computer system according to an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention more clear, the technical solutions of 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. It will be apparent that the described embodiments are some, but not all, embodiments of the invention. The components of the embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations.

Thus, the following detailed description of the embodiments of the invention, as presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. 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.

It should be noted that: like reference numerals and letters denote like items in the following figures, and thus once an item is defined in one figure, no further definition or explanation thereof is necessary in the following figures.

Furthermore, the terms "first," "second," and the like, are used merely to distinguish between descriptions and should not be construed as indicating or implying relative importance.

In the description of the present invention, it should also be noted that, unless explicitly specified and limited otherwise, terms such as "disposed," "connected," and the like are to be construed broadly, and for example, "connected" may be either fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium, and can be communication between two elements. The specific meaning of the above terms in the present invention can be understood by those of ordinary skill in the art according to the specific circumstances.

The following describes specific embodiments of the present invention in detail with reference to the drawings.

FIG. 1 is an interactive schematic diagram of an artificial intelligence based drive detection processing system according to an embodiment of the present disclosure. The artificial intelligence based drive detection processing system may include a computer system 100 and a plurality of industrial devices 200 communicatively coupled to the computer system 100. The artificial intelligence based drive detection processing system shown in fig. 1 is only one possible example, and in other possible embodiments, the artificial intelligence based drive detection processing system may include only a portion of the components shown in fig. 1 or may include other components as well.

In this embodiment, the computer system 100 and the plurality of industrial devices 200 in the driving detection processing system based on the artificial intelligence may cooperate to execute the driving detection processing method based on the artificial intelligence described in the following method embodiments, and the detailed description of the method embodiments may be referred to for the execution steps of the computer system 100 and the plurality of industrial devices 200.

In order to solve the foregoing technical problems in the background art, fig. 2 is a flowchart of an artificial intelligence-based driving detection processing method according to an embodiment of the present disclosure, and the artificial intelligence-based driving detection processing method according to the embodiment may be executed by the computer system 100 shown in fig. 1, and the detailed description of the artificial intelligence-based driving detection processing method is provided below.

In step 201, in response to the system control instruction, a label-free system operation feature vector matched with the target industrial equipment is obtained.

In the embodiment of the present invention, the system control command may be input by a user from the outside through a KVM (Keyboard Video Mouse, abbreviated as keyboard, video and mouse), and the system control command may be used to instruct to control the switching of the operation state of one target industrial device, or instruct to control a plurality of linked target industrial devices, which is not limited herein. In the embodiment of the present invention, the industrial device 200 may refer to industrial computers of different types and different systems, which is not limited herein. In the embodiment of the invention, the operation feature vector of the label-free system can refer to the relevant parameters of the system which is not labeled yet in each operation state, and optionally, the operation states can comprise operations such as starting, stopping, switching and the like.

Step 202, determining different system operation environments corresponding to the label-free system operation feature vectors matched with the target industrial equipment.

It should be understood that the problems solved by different industrial devices 200 are different, or the execution of a plurality of industrial devices 200 that cooperate to solve the same technical problem are different, based on which different system operation environments, i.e. different operation scenarios, respectively corresponding to the unlabeled system operation feature vectors that match the target industrial device can be determined.

Step 203, determining the HDMI state matching the different system operating environments based on the different system operating environments.

In the embodiment of the present invention, the HDMI state may refer to a state of an interface between the computer system 100 and the industrial equipment 200, and a specific way to change the HDMI state may be to call a SetChannel interface function to switch.

And 204, performing feature enhancement processing on the unlabeled system operation feature vectors matched with the target industrial equipment based on the HDMI state to form an initial system operation feature vector set matched with the corresponding system operation environment.

In order to enable the unlabeled system operation feature vector to have more features in the subsequent training process, the HDMI state may be added to the unlabeled system operation feature vector for feature enhancement processing, so as to obtain an initial system operation feature vector set that may be used as an input.

And step 205, calling a system control instruction generation model for countermeasure training through a historical system running state training sample, and generating control instructions for the initial system running feature vector set through the system control instruction generation model to obtain the control instructions of the target industrial equipment.

The system control instruction generation model comprises a feature extraction layer, a feature selection layer and an information filtering layer.

In the embodiment of the application, the pre-trained system control instruction generation model can be obtained by countertraining of a historical system running state training sample, and after the initial system running feature vector set is input, the control instruction of the target industrial equipment can be correspondingly obtained.

And 206, outputting a control instruction of the target industrial equipment.

The control instruction of the target industrial equipment obtained through the steps can accurately control the start, stop and operation of the target industrial equipment, the output control instruction of the target industrial equipment can be directly sent to the target industrial equipment for execution, and can be sent to the target industrial equipment after subsequent processing, the control instruction is not limited, the whole process is less in manual participation, and the control efficiency and the instruction accuracy of the industrial equipment 200 are improved in the application scene of the industrial equipment 200.

On the foregoing basis, as an alternative embodiment, the following embodiment is provided as an embodiment of the present invention.

Step 301, obtaining characteristics of a target industrial equipment set and historical parameters of a system start-up environment.

Step 302, obtaining a system operation feature vector set matched with a system control instruction generation model according to the features of the target industrial equipment set and the historical parameters of the system starting environment.

The system operation feature vector set comprises sample features of different historical system operation states.

And 303, determining the use environment of the system control instruction generation model, and determining sliding windows with different lengths according to the use environment of the system control instruction generation model.

Step 304, determining time node data carried by different sample data in the system operation feature vector set.

Step 305, processing the time node data carried by different sample data in the system operation feature vector set according to a preset iterative algorithm matched with the system control instruction generation model and sliding windows with different lengths, and determining sample system control data sets of different historical system operation states matched with the system control instruction generation model.

Wherein the sample system control data set comprises at least one set of sample system control data.

And step 306, training a system control instruction generation model according to the sample system control data set, and determining the super parameters of the feature extraction layer in the system control instruction generation model.

Step 307, training the system control instruction generation model according to the sample system control data set, and determining the hyper-parameters of the feature selection layer in the system control instruction generation model.

Step 308, training the system control instruction generation model according to the sample system control data set, and determining the super parameters of the information filtering layer in the system control instruction generation model so as to realize the prediction of the instruction of the target industrial equipment in the system starting environment through the system control instruction generation model.

On this basis, in order to more clearly describe the solution provided by the embodiments of the present invention, the foregoing step 306 may be implemented by the following steps.

In a substep 306-1, a feature extraction layer in the model is generated from the system control instructions, and the sample system control data set is processed to determine initial parameters of the feature extraction layer.

In response to the initial parameters of the feature extraction layer, the sample system control dataset is processed by the feature extraction layer to determine iterative parameters of the feature extraction layer, substep 206-2.

In a substep 306-3, a loss function corresponding to the feature extraction layer is determined.

And 306-4, carrying out iterative updating on the parameters of the feature extraction layer according to the iterative parameters of the feature extraction layer.

Substep 306-5 is performed until the loss function of the feature extraction layer reaches a corresponding preset convergence threshold, and based on parameters in the feature extraction layer, feature selection vectors for each sample in the sample system control dataset can be extracted.

In order to more clearly describe the solution provided by the present invention on the basis of the foregoing, the foregoing step 307 may be implemented by the following embodiments.

In substep 307-1, the feature selection layer in the model is generated by the system control instructions, and the sample system control data set is processed to determine initial parameters of the feature selection layer.

Substep 307-2 substitutes different sample system control data in the sample system control data set into the loss function corresponding to the feature selection layer.

In substep 307-3, the corresponding iteration parameters of the feature selection layer are determined when the loss function meets the corresponding preset convergence threshold.

Substep 307-4, a loss function corresponding to the feature selection layer is determined.

And a substep 307-5 of iteratively updating the parameters of the feature selection layer according to the iteration parameters of the feature selection layer.

Substep 307-6 is performed until the loss function of the feature selection layer reaches a corresponding preset convergence threshold, and the instruction output results of different samples are determined based on the corresponding sample tags and the feature selection vector of each sample.

In the embodiment of the present invention, in order to more clearly describe the solution provided by the present invention, the foregoing step 308 may be implemented by the following steps.

In response to the initial parameters of the information filtering layer, the sample system control dataset is processed through the information filtering layer to determine iteration parameters of the information filtering layer, substep 308-1.

In sub-step 308-2, according to the iteration parameters of the information filtering layer, the parameters of the information filtering layer are iteratively updated by the sample system control data set, so as to judge the running states of the historical system corresponding to different samples in the sample set.

In addition to the foregoing, as an alternative specific implementation, the embodiment of the present invention further includes the following implementation.

Step 401, substituting the sample system control data set with different sample system control data set into the loss function corresponding to the information filtering layer.

Step 402, determining that the information filter layer corresponds to the iteration parameter when the loss function meets the corresponding preset convergence threshold.

Step 403, determining a loss function corresponding to the information filtering layer.

And step 404, carrying out iterative updating on the parameters of the information filter layer according to the iterative parameters of the information filter layer.

Step 405, until the loss function of the information filtering layer reaches the corresponding preset convergence threshold, and based on the parameters in the information filtering layer, the historical system running states corresponding to different samples in the sample set can be determined.

In order to more clearly describe the solution provided by the present invention, the embodiments of the present invention further include, for example, the following embodiments.

(1) And processing time node data carried by different sample data in the system operation feature vector set according to a preset iterative algorithm matched with the system control instruction generation model and sliding windows with different lengths, and determining test sample sets of different historical system operation states matched with the system control instruction generation model.

(2) And processing different test samples in the test sample set through the system control instruction generation model so as to realize testing of instruction output results output by the system control instruction generation model through the test sample set.

In addition to the above, in order to improve the security in controlling the industrial equipment 200, the computer system 100 is also communicatively connected to both the authentication server and the user login system, and the following embodiments are provided in the embodiments of the present invention before executing step 201.

Step 501, user operation data uploaded by the verification server is obtained, and security detection is performed on the system control instruction stored in the control instruction temporary storage space.

The control instruction temporary storage space comprises a target industrial equipment control instruction set, the target industrial equipment control instruction set comprises at least one system control instruction, and different system control instructions are respectively generated by different user login systems.

Step 502, if a first system control instruction in at least one system control instruction passes the security detection and the first system control instruction is a system control instruction with the most effective reference value in the target industrial equipment control instruction set, acquiring a hash parameter corresponding to the first system control instruction, acquiring an encrypted fingerprint carried by user operation data, and acquiring a first key corresponding to the verification server.

In step 503, the encrypted fingerprint is decrypted based on the first key, so as to obtain first encrypted information corresponding to the encrypted fingerprint.

And step 504, calculating the user operation data based on the password hash function to obtain second encryption information corresponding to the user operation data.

In step 505, if the first encrypted information is identical to the second encrypted information, the user operation data is authenticated, and an operation data vector is generated based on the authenticated user operation data.

Step 506, generating a vector label according to the hash parameter, generating a second system control instruction according to the vector label and the operation data vector, and adding the second system control instruction to the target industrial equipment control instruction set to obtain an updated control instruction temporary storage space.

Step 507, traversing the second system control instruction in the user login system cluster, so that other user login systems except the user login system generating the second system control instruction in the user login system cluster can respectively cache the second system control instruction in the temporary storage space.

Step 508, the number of system control instructions contained in the updated control instruction temporary storage space is obtained, the user login system corresponding to each system control instruction in the updated control instruction temporary storage space is determined, and the credibility weight matched with the user login system is obtained.

Step 509, updating the confidence threshold value corresponding to each system control instruction in the updated control instruction temporary storage space based on the number and the confidence weight of the system control instructions contained in the updated control instruction temporary storage space.

And 510, marking the updated system control instruction with the confidence coefficient threshold value larger than the preset threshold value with an instruction valid tag, and obtaining the current confidence coefficient corresponding to the system control instruction with the instruction valid tag.

In step 511, if the difference between the current confidence level and the confidence level corresponding to the most effective reference value in the control instruction safety space is within the preset effective range, the system control instruction with the instruction effective tag is added to the control instruction safety space.

And step 512, if the difference between the current confidence coefficient and the confidence coefficient corresponding to the most effective reference value in the control instruction safety space is outside the preset effective range, updating the confidence coefficient of the system control instruction with the instruction effective label, and adding the updated system control instruction with the instruction effective label into the control instruction safety space.

The control instruction safety space is used for storing all system control instructions marked with instruction valid tags.

On the basis of the foregoing, the control instruction temporary storage space includes a plurality of system control instruction sets, and the plurality of system control instruction sets includes a target industrial equipment control instruction set, and as an alternative specific embodiment, the foregoing step 501 may be implemented and executed in the following manner.

In the sub-step 501-1, a plurality of system control instruction sets are obtained from the control instruction temporary storage space, and the initial system control instruction numbers corresponding to the plurality of system control instruction sets respectively are obtained.

In sub-step 501-2, the plurality of system control instruction sets are ordered based on the initial number of system control instructions, and the security detection is sequentially performed on the system control instructions included in each system control instruction set based on the ordering order of each system control instruction set.

On the basis of the foregoing, as an alternative embodiment, the foregoing step 502 may be implemented by performing the following steps.

In the substep 502-1, if there are target industrial equipment control instruction sets in which the system control instructions pass the security detection in the multiple system control instruction sets, the system control instruction with the most effective reference value in the target industrial equipment control instruction set is used as a first system control instruction, and a hash parameter corresponding to the first system control instruction is obtained.

In order to more clearly describe the solution provided by the present invention, the embodiments of the present invention also provide, for example, the following embodiments.

(1) If the system control instruction which does not pass through the safety detection exists in the target industrial equipment control instruction set, and the first system control instruction is the system control instruction with the most effective reference value in the system control instruction which passes through the safety detection in the target industrial equipment control instruction set, acquiring a hash parameter corresponding to the first system control instruction, and generating a second system control instruction according to the user operation data and the hash parameter.

(2) And forming a new system control instruction set by all system control instructions passing through safety detection in the target industrial equipment control instruction set and the second system control instruction, and determining the new system control instruction set and the target industrial equipment control instruction set as updated control instruction temporary storage space.

In addition, as an alternative embodiment, the embodiment of the present invention includes, for example, the following embodiment.

(1) If the system control instructions which do not pass through the safety detection exist in the plurality of system control instruction sets, respectively counting the target number of the system control instructions which pass through the safety detection in each system control instruction set, and determining the system control instruction set with the largest target number as a target industrial equipment control instruction set.

(2) And acquiring a system control instruction with the most effective reference value from system control instructions which are contained in the target industrial equipment control instruction set and pass through safety detection as a first system control instruction, and acquiring a hash parameter corresponding to the first system control instruction.

(3) Generating a second system control instruction according to the user operation data and the hash parameters, and forming a new system control instruction set by all the system control instructions passing through the safety detection in the target industrial equipment control instruction set and the second system control instruction.

(4) The new system control instruction set and the plurality of system control instruction sets are determined as updated control instruction scratch pad space.

On the basis of the foregoing, as an alternative embodiment, the foregoing step 509 may be performed, for example, in the following manner.

In the sub-step 509-1, the number of the instructions corresponding to the updated target industrial equipment control instruction set and the remaining system control instruction set is obtained from the updated control instruction temporary storage space, and the occurrence frequency of each system control instruction in the updated target industrial equipment control instruction set and the remaining system control instruction set is counted respectively.

In step 509-2, based on the number of instructions and the occurrence frequency, the confidence threshold corresponding to each system control instruction in the updated control instruction temporary storage space is reckoned.

The embodiment of the invention provides an artificial intelligence based drive detection processing cloud platform 110, which is applied to a computer system 100, wherein the computer system 100 is in communication connection with a plurality of industrial devices 200, as shown in fig. 3, the artificial intelligence based drive detection processing cloud platform 110 comprises:

and the acquisition module 1101 is configured to acquire, in response to the system control instruction, a label-free system operation feature vector matched with the target industrial device.

A determining module 1102, configured to determine different system operating environments corresponding to the unlabeled system operation feature vectors that are matched with the target industrial device respectively; determining HDMI states matched with different system operating environments based on the different system operating environments; and performing feature enhancement processing on the unlabeled system operation feature vectors matched with the target industrial equipment based on the HDMI state to form an initial system operation feature vector set matched with the corresponding system operation environment.

An output module 1103, configured to invoke a system control instruction generation model for countertraining via a historical system running state training sample, and generate a control instruction for an initial system running feature vector set through the system control instruction generation model, so as to obtain a control instruction of a target industrial device, where the system control instruction generation model includes a feature extraction layer, a feature selection layer, and an information filtering layer; and outputting a control instruction of the target industrial equipment.

Further, the determining module 1102 is further configured to:

acquiring characteristics of a target industrial equipment set and historical parameters of a system starting environment; obtaining a system operation feature vector set matched with a system control instruction generation model according to the features of the target industrial equipment set and the historical parameters of the system starting environment, wherein the system operation feature vector set comprises sample features of different historical system operation states; determining the use environment of a system control instruction generation model, and determining sliding windows with different lengths according to the use environment of the system control instruction generation model; determining time node data carried by different sample data in a system operation feature vector set; according to a preset iterative algorithm matched with a system control instruction generation model and sliding windows with different lengths, processing time node data carried by different sample data in a system operation feature vector set, and determining sample system control data sets of different historical system operation states matched with the system control instruction generation model, wherein the sample system control data sets comprise at least one group of sample system control data; training a system control instruction generation model according to a sample system control data set, and determining super parameters of a feature extraction layer in the system control instruction generation model; training a system control instruction generation model according to a sample system control data set, and determining hyper-parameters of a feature selection layer in the system control instruction generation model; training a system control instruction generation model according to the sample system control data set, and determining hyper-parameters of an information filtering layer in the system control instruction generation model so as to predict the instruction of the target industrial equipment in the system starting environment through the system control instruction generation model.

Further, the determining module 1102 is specifically configured to:

generating a feature extraction layer in the model through a system control instruction, and processing a sample system control data set to determine initial parameters of the feature extraction layer; responding to the initial parameters of the feature extraction layer, processing a sample system control data set through the feature extraction layer, and determining iteration parameters of the feature extraction layer; determining a loss function corresponding to the feature extraction layer; according to the iteration parameters of the feature extraction layer, carrying out iteration update on the parameters of the feature extraction layer; until the loss function of the feature extraction layer reaches a corresponding preset convergence threshold, and based on parameters in the feature extraction layer, feature selection vectors of each sample in the sample system control data set can be extracted.

Further, the determining module 1102 is specifically configured to:

generating a feature selection layer in the model through a system control instruction, and processing a sample system control data set to determine initial parameters of the feature selection layer; substituting different sample system control data in the sample system control data set into a loss function corresponding to the feature selection layer; determining corresponding iteration parameters of the feature selection layer when the loss function meets a corresponding preset convergence threshold; determining a loss function corresponding to the feature selection layer; according to the iteration parameters of the feature selection layer, carrying out iteration update on the parameters of the feature selection layer; and determining instruction output results of different samples based on the corresponding sample labels and the feature selection vector of each sample until the loss function of the feature selection layer reaches a corresponding preset convergence threshold.

Further, the determining module 1102 is specifically configured to:

responding to the initial parameters of the information filtering layer, processing the sample system control data set through the information filtering layer, and determining the iteration parameters of the information filtering layer; according to the iteration parameters of the information filtering layer, the parameters of the information filtering layer are iteratively updated through the sample system control data set, so that the historical system running states corresponding to different samples in the sample set are judged.

Further, the determining module 1102 is specifically further configured to:

substituting different sample system control data in the sample system control data set into a loss function corresponding to the information filtering layer; determining corresponding iteration parameters of the information filter layer when the loss function meets a corresponding preset convergence threshold; determining a loss function corresponding to the information filtering layer; according to the iteration parameters of the information filter layer, carrying out iteration update on the parameters of the information filter layer; and judging the historical system running states corresponding to different samples in the sample set based on the parameters in the information filter layer until the loss function of the information filter layer reaches the corresponding preset convergence threshold.

Further, the computer system 100 is also communicatively connected to both the authentication server and the user login system, and the determining module 1102 is further configured to:

The method comprises the steps of obtaining user operation data uploaded by a verification server, and carrying out safety detection on system control instructions stored in a control instruction temporary storage space, wherein the control instruction temporary storage space comprises a target industrial equipment control instruction set, the target industrial equipment control instruction set comprises at least one system control instruction, and different system control instructions are respectively generated by different user login systems; if a first system control instruction in at least one system control instruction passes the safety detection and the first system control instruction is the system control instruction with the most effective reference value in the target industrial equipment control instruction set, acquiring a hash parameter corresponding to the first system control instruction, acquiring an encrypted fingerprint carried by user operation data, and acquiring a first key corresponding to the verification server; decrypting the encrypted fingerprint based on the first key to obtain first encrypted information corresponding to the encrypted fingerprint; operating the user operation data based on the password hash function to obtain second encryption information corresponding to the user operation data; if the first encryption information is the same as the second encryption information, the user operation data is verified and passed, and an operation data vector is generated based on the verified user operation data; generating a vector label according to the hash parameter, generating a second system control instruction according to the vector label and the operation data vector, and adding the second system control instruction to a target industrial equipment control instruction set to obtain an updated control instruction temporary storage space; traversing the second system control instruction in the user login system cluster to enable other user login systems except the user login system generating the second system control instruction in the user login system cluster, and respectively caching the second system control instruction into the belonging temporary storage space; acquiring the number of system control instructions contained in the updated control instruction temporary storage space, determining a user login system corresponding to each system control instruction in the updated control instruction temporary storage space, and acquiring credibility weights matched with the user login system; updating the confidence threshold value corresponding to each system control instruction in the updated control instruction temporary storage space based on the number and the credibility weight of the system control instructions contained in the updated control instruction temporary storage space; marking an instruction valid tag on the system control instruction with the updated confidence coefficient threshold value larger than the preset threshold value, and obtaining the current confidence coefficient corresponding to the system control instruction with the instruction valid tag; if the difference between the current confidence coefficient and the confidence coefficient corresponding to the most effective reference value in the control instruction safety space is within a preset effective range, adding a system control instruction with an instruction effective label into the control instruction safety space; if the difference between the current confidence coefficient and the confidence coefficient corresponding to the most effective reference value in the control instruction safety space is outside a preset effective range, performing confidence coefficient updating on the system control instruction with the instruction effective label, and adding the updated system control instruction with the instruction effective label into the control instruction safety space, wherein the control instruction safety space is used for storing all the system control instructions marked with the instruction effective label.

Further, the control instruction temporary storage space includes a plurality of system control instruction sets, the plurality of system control instruction sets includes a target industrial equipment control instruction set, and the determining module 1102 is specifically configured to:

acquiring a plurality of system control instruction sets from a control instruction temporary storage space, and acquiring initial system control instruction numbers corresponding to the plurality of system control instruction sets respectively; the method comprises the steps of sequencing a plurality of system control instruction sets based on the initial system control instruction number, and sequentially carrying out safety detection on the system control instructions contained in each system control instruction set based on the sequencing order of each system control instruction set.

Further, the determining module 1102 is specifically configured to:

if a target industrial equipment control instruction set with the system control instructions passing through the safety detection exists in the plurality of system control instruction sets, taking the system control instruction with the most effective reference value in the target industrial equipment control instruction set as a first system control instruction, and acquiring a hash parameter corresponding to the first system control instruction.

It should be noted that, the implementation principle of the driving detection processing cloud platform 110 based on artificial intelligence may refer to the implementation principle of the driving detection processing method based on artificial intelligence, which is not described herein. It should be understood that the division of the modules of the above cloud platform is merely a division of logic functions, and may be fully or partially integrated into one physical entity or may be physically separated when actually implemented. And these modules may all be implemented in software in the form of calls by the processing element; or can be realized in hardware; the method can also be realized in a form of calling software by a processing element, and the method can be realized in a form of hardware by a part of modules. For example, the artificial intelligence-based drive detection processing cloud platform 110 may be a processing element that is set up separately, or may be implemented in a chip of the cloud platform, or may be stored in a memory of the cloud platform in a form of program codes, and the functions of the acquisition module 1101 may be invoked and executed by a processing element of the cloud platform. The implementation of the other modules is similar. In addition, all or part of the modules can be integrated together or can be independently implemented. The processing element described herein may be an integrated circuit having signal processing capabilities. In implementation, each step of the above method or each module above may be implemented by an integrated logic circuit of hardware in a processor element or an instruction in a software form.

For example, the modules above may be one or more integrated circuits configured to implement the methods above, such as: one or more specific integrated circuits (application specific integrated circuit, ASIC), or one or more microprocessors (digital signal processor, DSP), or one or more field programmable gate arrays (field programmable gate array, FPGA), or the like. For another example, when a module above is implemented in the form of a processing element scheduler code, the processing element may be a general purpose processor, such as a central processing unit (centralprocessing unit, CPU) or other processor that may invoke the program code. For another example, the modules may be integrated together and implemented in the form of a system-on-a-chip (SOC).

The embodiment of the invention provides a computer system 100, where the computer system 100 includes a processor and a nonvolatile memory storing computer instructions, and when the computer instructions are executed by the processor, the computer system 100 executes the driving detection processing cloud platform 110 based on artificial intelligence. As shown in fig. 4, fig. 4 is a block diagram illustrating a computer system 100 according to an embodiment of the present invention. The computer system 100 includes an artificial intelligence based drive detection processing cloud platform 110, a memory 111, a processor 112, and a communication unit 113.

For data transmission or interaction, the memory 111, the processor 112 and the communication unit 113 are electrically connected to each other directly or indirectly. For example, the elements may be electrically connected to each other via one or more communication buses or signal lines. The artificial intelligence based drive detection processing cloud platform 110 includes at least one software functional module that may be stored in the memory 111 in the form of software or firmware (firmware) or cured in an Operating System (OS) of the computer system 100. The processor 112 is configured to execute the artificial intelligence-based driving detection processing cloud platform 110 stored in the memory 111, for example, a software function module and a computer program included in the artificial intelligence-based driving detection processing cloud platform 110.

An embodiment of the present invention provides a readable storage medium, where the readable storage medium includes a computer program, and when the computer program runs, controls a computer system 100 where the readable storage medium is located to execute the driving detection processing method based on artificial intelligence.

In summary, the embodiment of the invention provides a driving detection processing method based on artificial intelligence, which is used for acquiring a label-free system operation feature vector matched with a target industrial system by responding to a system control instruction; determining different system operation environments corresponding to the operation feature vectors of the label-free system matched with the target industrial system respectively; further, based on different system operation environments, determining HDMI states matched with the different system operation environments; then, based on the HDMI state, carrying out feature enhancement processing on the unlabeled system operation feature vectors matched with the target industrial system to form an initial system operation feature vector set matched with the corresponding system operation environment; then, a system control instruction generation model for countertraining by using a historical system running state training sample is called, and control instruction generation is carried out on an initial system running feature vector set through the system control instruction generation model to obtain a control instruction of a target industrial system, wherein the system control instruction generation model comprises a feature extraction layer, a feature selection layer and an information filtering layer; finally, a control instruction of the target industrial system is output, and through the steps, the automatic control of the industrial system is realized by skillfully utilizing a system control instruction generation model.

The foregoing description, for purpose of explanation, has been presented with reference to particular embodiments. However, the illustrative discussions above are not intended to be exhaustive or to limit the disclosure to the precise forms disclosed. Many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the disclosure and its practical application, to thereby enable others skilled in the art to best utilize the disclosure and various embodiments with various modifications as are suited to the particular use contemplated. The foregoing description, for purpose of explanation, has been presented with reference to particular embodiments. However, the illustrative discussions above are not intended to be exhaustive or to limit the disclosure to the precise forms disclosed. Many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the disclosure and its practical application, to thereby enable others skilled in the art to best utilize the disclosure and various embodiments with various modifications as are suited to the particular use contemplated.

Claims

1. A method for driving detection processing based on artificial intelligence, which is applied to a computer system, wherein the computer system is in communication connection with a plurality of industrial systems, and the method comprises:

determining, based on the different system operating environments, HDMI states that match the different system operating environments;

based on the HDMI state, performing feature enhancement processing on the unlabeled system operation feature vector matched with the target industrial system to form an initial system operation feature vector set matched with a corresponding system operation environment;

invoking a system control instruction generation model for countermeasure training through a historical system running state training sample, and generating a control instruction for the initial system running feature vector set through the system control instruction generation model to obtain a control instruction of the target industrial system, wherein the system control instruction generation model comprises a feature extraction layer, a feature selection layer and an information filtering layer;

and outputting a control instruction of the target industrial system.

2. The method according to claim 1, wherein the method further comprises:

obtaining a system operation feature vector set matched with the system control instruction generation model according to the features of the target industrial system set and the historical parameters of a system starting environment, wherein the system operation feature vector set comprises sample features of different historical system operation states;

determining the use environment of the system control instruction generation model, and determining sliding windows with different lengths according to the use environment of the system control instruction generation model;

determining time node data carried by different sample data in the system operation feature vector set;

processing time node data carried by different sample data in a system operation feature vector set according to a preset iterative algorithm matched with the system control instruction generation model and sliding windows with different lengths, and determining sample system control data sets of different historical system operation states matched with the system control instruction generation model, wherein the sample system control data sets comprise at least one group of sample system control data;

Training the system control instruction generation model according to the sample system control data set, and determining the hyper-parameters of a feature extraction layer in the system control instruction generation model;

training the system control instruction generation model according to the sample system control data set, and determining the hyper-parameters of a feature selection layer in the system control instruction generation model;

training the system control instruction generation model according to the sample system control data set, and determining the hyper-parameters of an information filtering layer in the system control instruction generation model so as to realize the prediction of the instruction of the target industrial system in the system starting environment through the system control instruction generation model.

3. The method of claim 2, wherein training the system control instruction generation model based on the sample system control data set, determining hyper-parameters of a feature extraction layer in the system control instruction generation model, comprises:

generating a feature extraction layer in a model through the system control instruction, and processing the sample system control data set to determine initial parameters of the feature extraction layer;

Processing the sample system control data set by the feature extraction layer in response to initial parameters of the feature extraction layer, determining iterative parameters of the feature extraction layer;

determining a loss function corresponding to the feature extraction layer;

and extracting the feature selection vector of each sample in the sample system control data set based on the parameters in the feature extraction layer until the loss function of the feature extraction layer reaches a corresponding preset convergence threshold.

4. A method according to claim 3, wherein said training the system control instruction generation model from the sample system control data set to determine hyper-parameters of feature selection layers in the system control instruction generation model comprises:

generating a feature selection layer in a model through the system control instruction, and processing the sample system control data set to determine initial parameters of the feature selection layer;

Determining that the feature selection layer corresponds to the iteration parameter when the loss function meets a corresponding preset convergence threshold;

determining a loss function corresponding to the feature selection layer;

5. The method of claim 2, wherein training the system control instruction generation model based on the sample system control data set, determining hyper-parameters of an information filter layer in the system control instruction generation model, comprises:

and according to the iteration parameters of the information filtering layer, carrying out iteration update on the parameters of the information filtering layer through the sample system control data set so as to judge the historical system running states corresponding to different samples in the sample set.

6. The method of claim 5, wherein the method further comprises:

substituting different sample system control data in the sample system control data set into a loss function corresponding to the information filter layer;

determining that the information filter layer corresponds to the iteration parameter when the loss function meets a corresponding preset convergence threshold;

determining a loss function corresponding to the information filtering layer;

according to the iteration parameters of the information filtering layer, carrying out iteration update on the parameters of the information filtering layer;

and judging the historical system running states corresponding to different samples in the sample set based on the parameters in the information filter layer until the loss function of the information filter layer reaches a corresponding preset convergence threshold.

7. The method of claim 1, wherein the computer system is further communicatively coupled to both the authentication server and the user login system, the method comprising, prior to the obtaining of the unlabeled system operation feature vector matching the target industrial system in response to the system control command:

If a first system control instruction in the at least one system control instruction passes through security detection and is a system control instruction with the most effective reference value in the target industrial system control instruction set, acquiring a hash parameter corresponding to the first system control instruction, acquiring an encrypted fingerprint carried by the user operation data, and acquiring a first key corresponding to the verification server;

operating the user operation data based on a password hash function to obtain second encryption information corresponding to the user operation data;

if the first encryption information is the same as the second encryption information, the user operation data is verified to pass, and an operation data vector is generated based on the verified user operation data;

generating a vector label according to the hash parameter, generating a second system control instruction according to the vector label and the operation data vector, and adding the second system control instruction to the target industrial system control instruction set to obtain an updated control instruction temporary storage space;

Traversing the second system control instruction in a user login system cluster to enable other user login systems except the user login system generating the second system control instruction in the user login system cluster to respectively cache the second system control instruction into the temporary storage space;

acquiring the number of system control instructions contained in the updated control instruction temporary storage space, determining a user login system corresponding to each system control instruction in the updated control instruction temporary storage space, and acquiring credibility weights matched with the user login systems;

updating the confidence threshold value corresponding to each system control instruction in the updated control instruction temporary storage space based on the number of the system control instructions contained in the updated control instruction temporary storage space and the credibility weight;

if the difference between the current confidence coefficient and the confidence coefficient corresponding to the most effective reference value in the control instruction safety space is within a preset effective range, adding the system control instruction with the instruction effective label into the control instruction safety space;

8. The method of claim 7, wherein the control instruction scratch space comprises a plurality of sets of system control instructions, the plurality of sets of system control instructions comprising the set of target industrial system control instructions;

the safety detection of the system control instruction stored in the control instruction temporary storage space comprises the following steps:

acquiring the plurality of system control instruction sets from the control instruction temporary storage space, and acquiring initial system control instruction numbers respectively corresponding to the plurality of system control instruction sets;

and sequencing the plurality of system control instruction sets based on the initial system control instruction number, and sequentially performing security detection on the system control instructions contained in each system control instruction set based on the sequencing sequence of each system control instruction set.

9. The method of claim 8, wherein if a first system control command of the at least one system control command passes the security detection and the first system control command is a system control command having a most effective reference value in the target industrial system control command set, obtaining a hash parameter corresponding to the first system control command comprises:

if a target industrial system control instruction set, in which the system control instructions pass through the safety detection, exists in the plurality of system control instruction sets, taking the system control instruction with the most effective reference value in the target industrial system control instruction set as the first system control instruction, and acquiring a hash parameter corresponding to the first system control instruction.

10. A drive detection processing cloud platform based on artificial intelligence, characterized in that the drive detection processing cloud platform is applied to a computer system, wherein the computer system is in communication connection with a plurality of industrial systems, and the cloud platform comprises:

the determining module is used for determining different system operation environments corresponding to the label-free system operation feature vectors matched with the target industrial system respectively; determining, based on the different system operating environments, HDMI states that match the different system operating environments; based on the HDMI state, performing feature enhancement processing on the unlabeled system operation feature vector matched with the target industrial system to form an initial system operation feature vector set matched with a corresponding system operation environment;