CN108460452A - A kind of self-desttruction equipment and method of artificial intelligence behavior body - Google Patents
A kind of self-desttruction equipment and method of artificial intelligence behavior body Download PDFInfo
- Publication number
- CN108460452A CN108460452A CN201810234182.7A CN201810234182A CN108460452A CN 108460452 A CN108460452 A CN 108460452A CN 201810234182 A CN201810234182 A CN 201810234182A CN 108460452 A CN108460452 A CN 108460452A
- Authority
- CN
- China
- Prior art keywords
- self
- destruction
- decision
- module
- artificial intelligence
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/008—Artificial life, i.e. computing arrangements simulating life based on physical entities controlled by simulated intelligence so as to replicate intelligent life forms, e.g. based on robots replicating pets or humans in their appearance or behaviour
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Robotics (AREA)
- Selective Calling Equipment (AREA)
Abstract
The invention discloses a kind of self-desttruction equipment and method of artificial intelligence behavior body, which includes：Sensing module, decision-making module, self-destruction module and receiving module；Wherein, sensing module is used to acquire the event information of artificial intelligence behavior body, and the event information of acquisition is sent to decision-making module；Decision-making module is used for the event information according to acquisition, judges whether to need to execute self-destruction；If it is, according to judging result, command-destruct is generated, and command-destruct is sent to self-destruction module；Self-destruction module is used to be instructed according to command-destruct or remote self-destruction, calculates each component and self-destruction flow, logical place and the physical location of service, and being destroyed according to self-destruction flow need to self-destructed target element, service code, data, file system or physical hardware.Using the embodiment of the present invention, the out of control of artificial intelligence behavior body can be prevented, ensures artificial intelligent behavior body not safety of entail dangers to mankind when being serviced for the mankind.
Description
Technical field
The present invention relates to the self-desttruction equipments and side of field of artificial intelligence more particularly to a kind of artificial intelligence behavior body
Method.
Background technology
Artificial intelligence behavior body, which refers to one kind, can perceive external environment and as input, be determined by internal algorithm
Plan, and interact using itself driving device and physical world the autonomous hardware entities of behavior.Wherein, external environment includes institute
Locate natural environment and organism etc.；Internal algorithm be preset inside artificial intelligence behavior body, the program for generating decision；It drives
Dynamic device refer to can interacting with physical world or changeable artificial intelligence behavior body residing for space coordinate hardware；What autonomy referred to
It is artificial intelligence behavior body oneself to set object function or make decisions on one's own.Artificial intelligence behavior body includes but not limited to autonomy
Robot (Intelligent Robot), autonomous driving vehicle (Self-driving Cars), artificial intelligence weapon
(Artificial Intelligence Weapon), artificial life (Artificial Life), intelligent agent
(Intelligent Agent), autonomous main body (Autonomous Agent).By taking autonomous robot soldier as an example, external environment
The characterizing population group exactly perceived by sensor, internal algorithm are exactly to start to attack after identifying features described above crowd, action dress
It is exactly machine leg and attack weapon to set, and such robot soldier can afield act on one's own, make decisions on one's own.
And the out of control of artificial intelligence behavior body should include three elements：The artificial intelligence of the forms such as robot has " behavior
Ability "；Artificial intelligence behavior body has " self evolvability "；Artificial intelligence behavior body is due to quilts such as design defect or loopholes
Hacker or extremely visitor utilize and change the design original intention of artificial intelligence product.In order to cope with safe prestige caused by artificial intelligence
The side of body, prevents the out of control of artificial intelligence, or even endanger the safety of human society, it is necessary to study a kind of artificial intelligence behavior body from
I destroys technology.
Self-destruction technology is all widely used in fields such as guided missile, unmanned plane, unattended equipment.Current self-destruction method kind
Class is various such as to destroy data by rewriting data, magnetic destruction, destruction key mode, or using high-tension circuit setting or mistake
Circuit is burnt in stream operation, and chip is destroyed using strong acid, ignites Micro-bomb demolition chip etc..And existing self-desttruction equipment is generally
Mechanical self-desttruction equipment and electronics self-desttruction equipment, general electronics self-desttruction equipment mainly use application-specific integrated circuit integrated circuit or micro energy lose
Microcontroller (microprocessor) self-desttruction equipment, controls guided missile using predetermined software program or electronic telecontrol technology, realizes
Self-destruction.Further, it is also possible to which the design parameter to software is set, real-time self-destruction or delay self-destruction are realized, however currently also
It is not directed to the self-desttruction equipment of AI behavior bodies, existing self-desttruction equipment cannot be satisfied safe shape of the user residing for AI behavior bodies
Whether state and the real-time automatic sensing AI behaviors body of behavior will be in runaway condition, and according to security factor auto-destruct AI behavior bodies
It is remote according to actual demand can not to meet user under specific condition for relevant functional unit, service module and physical system
The demand of the related functional components of AI behaviors, service module and physical system is destroyed in journey remote control.
Invention content
The embodiment of the present invention proposes a kind of self-desttruction equipment and method of artificial intelligence behavior body, can prevent artificial intelligence behavior
Body it is out of control, ensure artificial intelligent behavior body not safety of entail dangers to mankind when being serviced for the mankind.
The embodiment of the present invention provides a kind of self-desttruction equipment of artificial intelligence behavior body, including：Sensing module, decision-making module,
Self-destruction module and receiving module；
Wherein, the decision-making module is connect with the self-destruction module, sensing module respectively；The self-destruction module connects with described
Receive module connection；
The sensing module is used to acquire the event information of the artificial intelligence behavior body, and the event information of acquisition is sent out
Give the decision-making module；
The decision-making module is used for the event information according to the acquisition, judges whether to need to execute self-destruction；If it is,
According to judging result, command-destruct is generated, and the command-destruct is sent to self-destruction module；If it is not, then return and after
It is continuous to be judged；
The receiving module is used to receive the remote self-destruction instruction of remote equipment transmission；Wherein, the remote self-destruction instruction
With highest execution priority；
The self-destruction module is used to be instructed according to the command-destruct or remote self-destruction, calculate each component and service from
Flow, logical place and physical location are ruined, and being destroyed according to the self-destruction flow need to self-destructed target element, service code, number
According to, file system or physical hardware.
Further, the decision-making module is used for the event information according to the acquisition, judges whether to execute self-destruction, specifically
For：
The decision-making module carries out principal component analysis and feature extraction to the event information of the acquisition, after obtaining dimensionality reduction
Security attribute obtains the result of decision in conjunction with rough set theory, is judged whether to execute self-destruction according to the result of decision.
Further, the decision-making module carries out principal component analysis and feature extraction to the event information of the acquisition, obtains
The security attribute after dimensionality reduction is obtained, obtains the result of decision in conjunction with rough set theory, specially：
If the decision condition number of the artificial intelligence behavior body is r, collected security incident sum is N, Mei Gean
Total event has r dimensions, N number of vector to be expressed as：x1, x2,…,xn, then the safety condition attribute matrix X of n rows r row is constituted；
The decision-making module calculates the average vector and covariance matrix of the safety condition attribute matrix X, then calculates institute
The characteristic value and feature vector for stating covariance matrix are ranked up according to the characteristic value is descending, m master before calculated for rank
The contribution rate of accumulative total of member；The contribution rate of accumulative total is for the component after weighing dimensionality reduction to the fidelity of initial data；
Extract the feature vector corresponding to the preceding m pivot, constitute transformation matrix T, and by Y=XT calculate it is described before
The principal component of m pivot obtains the result of decision in conjunction with rough set theory.
Further, the self-destruction module includes self-destruction executive module and self-destruction control assembly；
The self-destruction executive module is used to be instructed according to the command-destruct or remote self-destruction, according to preset self-destruction component
Decision tree calculates each component and self-destruction flow, logical place and the physical location of service, and is destroyed according to the self-destruction flow
Need self-destructed target element, service code, data, file system or physical hardware；
The self-destruction control assembly is used to confirm the command-destruct or the credibility of remote self-destruction instruction, and monitors entire
Self-destruction flow is intervened when self-destruction is failed.
Further, the self-desttruction equipment is serially connected between power supply system and artificial intelligence decision system.
Correspondingly, the embodiment of the present invention also provides a kind of self-destruction method of artificial intelligence behavior body, including：
Acquire the event information of the artificial intelligence behavior body；
According to the event information of the acquisition, judge whether to need to execute self-destruction；If it is, according to judging result, it is raw
It is sent to self-destruction module at command-destruct, and by the command-destruct；If it is not, then returning and continuing to judge；
According to the command-destruct, each component and self-destruction flow, logical place and the physical location of service are calculated, and press
Being destroyed according to the self-destruction flow needs self-destructed target element, service code, data, file system or physical hardware.
Further, the self-destruction method further includes：
Receive the remote self-destruction instruction that remote equipment is sent；
It is instructed according to the remote self-destruction, judges whether the remote self-destruction instruction is credible, if the remote self-destruction refers to
It enables source credible, then calculates each component and self-destruction flow, logical place and the physical location of service, and according to the self-destruction stream
Journey, which is destroyed, needs self-destructed target element, service code, data, file system or physical hardware.
Further, the event information according to the acquisition judges whether to need to execute self-destruction, specially：To institute
The event information for stating acquisition carries out principal component analysis and feature extraction, obtains the security attribute after dimensionality reduction, is managed in conjunction with rough set
By the result of decision is obtained, judged whether to execute self-destruction according to the result of decision；
Wherein, the event information of the acquisition includes the status information of artificial intelligence behavior body itself.
Further, the event information to the acquisition carries out principal component analysis and feature extraction, after obtaining dimensionality reduction
Security attribute, obtain the result of decision in conjunction with rough set theory, specially：
If the decision condition number of the artificial intelligence behavior body is r, collected security incident sum is N, Mei Gean
Total event has r dimensions, N number of vector to be expressed as：x1, x2,…,xn, then the safety condition attribute matrix X of n rows r row is constituted；
The average vector and covariance matrix of the safety condition attribute matrix X are calculated, then calculates the covariance matrix
Characteristic value and feature vector, be ranked up according to the characteristic value is descending, the accumulative contribution of m pivot before calculated for rank
Rate；The contribution rate of accumulative total is for the component after weighing dimensionality reduction to the fidelity of initial data；
Extract the feature vector corresponding to the preceding m pivot, constitute transformation matrix T, and by Y=XT calculate it is described before
The principal component of m pivot obtains the result of decision in conjunction with rough set theory.
Further, self-destruction flow, logical place and the physical location for calculating each component and service, and according to institute
Stating the destruction of self-destruction flow needs self-destructed target element, service code, data, file system or physical hardware, specially：
According to the command-destruct each component and the self-destruction stream of service are calculated according to preset self-destruction component decision tree
Journey, logical place and physical location, and being destroyed according to the self-destruction flow need to self-destructed target element, service code, data, text
Part system or physical hardware.
Implement the embodiment of the present invention, has the advantages that：
The self-desttruction equipment and method of a kind of artificial intelligence behavior body provided in an embodiment of the present invention, can the artificial intelligence of automatic collection
The event information of energy behavior body, and judge whether it is in runaway condition, if it needs to execute self-destruction, if it is, by certainly
It ruins module and self-destruction is carried out to each component, service and system, realize self pin of multi-level, fine-grained artificial intelligence behavior body
It ruins.In addition, self-desttruction equipment can also receive the remote self-destruction instruction of remote equipment transmission, practicability and safety are improved.
Description of the drawings
Fig. 1 is a kind of structural schematic diagram of embodiment of the self-desttruction equipment of artificial intelligence behavior body provided by the invention.
Specific implementation mode
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation describes, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, those of ordinary skill in the art are obtained every other without creative efforts
Embodiment shall fall within the protection scope of the present invention.
It is a kind of structural representation of embodiment of the self-desttruction equipment of artificial intelligence behavior body provided by the invention referring to Fig. 1
Figure, the self-desttruction equipment include：Sensing module 1, decision-making module 2, self-destruction module 3 and receiving module 4.Wherein, decision-making module 2 is distinguished
It is connect with self-destruction module 3, sensing module 1；Self-destruction module 3 is connect with receiving module 4.
In the present embodiment, self-desttruction equipment is serially connected between power supply system and artificial intelligence decision system, can with but not
It is limited to realize the judgement of the reception and trigger condition of control instruction using micro energy lose chip.
Sensing module 1 is used to acquire the event information of artificial intelligence behavior body, and the event information of acquisition is sent to certainly
Plan module 2.
In the present embodiment, sensing module 1 passes through the sensor of artificial intelligence behavior body (such as sound, light, electricity, radar, red
Outside, the sensors such as gravity) environment around real-time perception, acquire the event such as letters such as target person, event and barrier of surrounding
Breath, the event information of record artificial intelligence system outwardly and inwardly, foundation is provided for self-destruction condition criterion.
Decision-making module 2 is used for the event information according to acquisition, judges whether to need to execute self-destruction；If it is, according to sentencing
Break as a result, generating command-destruct, and command-destruct is sent to self-destruction module 3；If it is not, then returning and continuing to sentence
It is disconnected.
In the present embodiment, decision-making module 2 is used for the event information according to acquisition, judges whether to execute self-destruction, specially：
Decision-making module carries out principal component analysis and feature extraction to the event information of acquisition, obtains the security attribute after dimensionality reduction, in conjunction with
Rough set theory obtains the result of decision, is judged whether to execute self-destruction according to the result of decision.2 integrated application principal component of decision-making module point
Analysis (PCA) is the feature extraction and Data Dimensionality Reduction the advantages of and rough set theory can carry out the advantage of knowledge reasoning, automatic to judge
Whether artificial intelligence behavior body is in runaway condition.
Further, decision-making module 2 carries out principal component analysis and feature extraction to the event information of acquisition, after obtaining dimensionality reduction
Security attribute, obtain the result of decision in conjunction with rough set theory, specially：
If the decision condition number of artificial intelligence behavior body is r, collected security incident sum is N, each safe thing
Part has r dimensions, N number of vector to be expressed as：x1, x2,…,xn, then the safety condition attribute matrix X of n rows r row is constituted；
Decision-making module 3 calculates the average vector and covariance matrix of safety condition attribute matrix X, then calculates covariance matrix
Characteristic value and feature vector, be ranked up according to characteristic value is descending, the contribution rate of accumulative total of m pivot before calculated for rank；
Contribution rate of accumulative total is for the component after weighing dimensionality reduction to the fidelity of initial data；m≤n；
Feature vector before extraction corresponding to m pivot constitutes transformation matrix T, and calculates preceding m pivot by Y=XT
Principal component, obtain the result of decision in conjunction with rough set theory.
In the present embodiment, the self-destruction condition automatic judging method based on rough set theory is specially：When security attribute drops
Become to be made of main variables Y after dimension, if K=(U, A, V, P) is a knowledge-representation system.Wherein, security incident record pair
It is conditional attribute set to answer object set U, A=C ∪ D, C, and D is decision attribute set, and V is the property value set of condition, and P is corresponding
Decision attribute values.Therefore, following decision table can be constructed according to knowledge-representation system, and the attribute of decision table is carried out brief：
In conjunction with upper table, if R is to destroy the derived classification of strategy set t, attribute set C indicates U/C to the division of set U.
Then R is represented by R=∪ E (R), and wherein E (R) indicates all knowledge by U/C expression of classifying.Finally according to rough set theory
Middle security attribute generates corresponding decision rule.
As a kind of citing of the present embodiment, decision-making module 2 can also pre-set boundary condition and security factor, once
Artificial intelligence behavior body has exceeded preset boundary condition and security factor, then automatic trigger self destroys program.
Receiving module 4 is used to receive the remote self-destruction instruction of remote equipment transmission.When receiving remote self-destruction instruction, answer
It is tested to command source, integrality and non repudiation with encrypted and digitally signed technology, it is ensured that the credibility of instruction.
Wherein, the self-destruction that remote self-destruction instruction and decision-making module generate, which is specified, should have highest priority, can interrupt
Any task that current manual's intelligent behavior is carrying out, if the self-destruction that remote self-destruction instruction and decision-making module generate is specified same
Shi Shengxiao should preferentially execute remote self-destruction instruction.
Self-destruction module 3 is used to be instructed according to command-destruct or remote self-destruction, calculate each component and service self-destruction flow,
Logical place and physical location, and being destroyed according to self-destruction flow need to self-destructed target element, service code, data, file system
Or physical hardware.
In the present embodiment, self-destruction module 3 includes self-destruction executive module and self-destruction control assembly.Self-destruction executive module is used for
It is instructed according to command-destruct or remote self-destruction, according to preset self-destruction component decision tree, calculates the self-destruction of each component and service
Flow, logical place and physical location, and being destroyed according to self-destruction flow need to self-destructed target element, service code, data, file
System or physical hardware；Self-destruction control assembly is used to confirm command-destruct or the credibility of remote self-destruction instruction, and monitors entire
Self-destruction flow is intervened when self-destruction is failed.
In the present embodiment, self-destruction module 3 can realize component self-destruction, service self-destruction, physical system self-destruction；Component self-destruction is
Some component for referring to auto-destruct artificial intelligence behavior body, such as robot arm assembly function, power supply module etc.；Service function is certainly
Ruin refers to the various services destroyed artificial intelligence behavior body and externally provided, such as voice service, Self-Service etc..Physical system self-destruction
Refer to destroy artificial intelligence behavior body physical system be allowed to be completely in unrecoverable state.
When self-destruction, system needs, according to the logical place and physical location of file where calculating each component and service, to carry out
Preliminary self-destruction destroys its code and instruction in memory, then to the storage file of the physical location of corresponding function component and
Data carry out erasing repeatedly and complete the complete self-destruction of irrecoverability.And for physical system self-destruction, then in a manner of irreversible
Destroy code, data and the file system of the decision system of artificial intelligence behavior body and to destroy the confession of artificial intelligence behavior body
Electric system is main target.Then it by control channel to auto-destruct chip, sends out from order is destroyed, is implanted into AI behavior bodies
The destruction unit of connection makes its energization that storage device and power-supply device destruction occur without artificial through decoder output high level
Operation carries out irreversible destruction to storage medium, AI systems is made to lose original function.
Correspondingly, the present invention provides a kind of self-destruction method of artificial intelligence behavior body, this approach includes the following steps：
Acquire the event information of artificial intelligence behavior body；
According to the event information of acquisition, judge whether to need to execute self-destruction；If it is, according to judging result, generate certainly
Instruction is ruined, and command-destruct is sent to self-destruction module；If it is not, then returning and continuing to judge；
According to command-destruct, each component and self-destruction flow, logical place and the physical location of service are calculated, and according to certainly
Ruining flow destruction needs self-destructed target element, service code, data, file system or physical hardware.
In the present embodiment, which further includes：Receive the remote self-destruction instruction that remote equipment is sent；According to long-range
Command-destruct, whether judgement remote self-destruction instruction is credible, if remote self-destruction command source is credible, calculates each component kimonos
Self-destruction flow, logical place and the physical location of business, and being destroyed according to the self-destruction flow need to self-destructed target element, service generation
Code, data, file system or physical hardware.
In the present embodiment, according to the event information of acquisition, judge whether to need to execute self-destruction, specially：To acquisition
Event information carries out principal component analysis and feature extraction, obtains the security attribute after dimensionality reduction, is obtained certainly in conjunction with rough set theory
Plan according to the result of decision as a result, judge whether to execute self-destruction；Wherein, the event information of acquisition includes artificial intelligence behavior body itself
Status information.
In the present embodiment, principal component analysis and feature extraction are carried out to the event information of acquisition, obtains the peace after dimensionality reduction
Full attribute obtains the result of decision, specially in conjunction with rough set theory：
If the decision condition number of artificial intelligence behavior body is r, collected security incident sum is N, each safe thing
Part has r dimensions, N number of vector to be expressed as：x1, x2,…,xn, then the safety condition attribute matrix X of n rows r row is constituted；
The average vector and covariance matrix of safety condition attribute matrix X are calculated, then calculates the characteristic value of covariance matrix
And feature vector, it is ranked up according to characteristic value is descending, the contribution rate of accumulative total of m pivot before calculated for rank；Accumulative contribution
Rate is for the component after weighing dimensionality reduction to the fidelity of initial data；
Feature vector before extraction corresponding to m pivot constitutes transformation matrix T, and calculates the preceding m by Y=XT
The principal component of pivot obtains the result of decision in conjunction with rough set theory.
In the present embodiment, each component and self-destruction flow, logical place and the physical location of service are calculated, and according to institute
Stating the destruction of self-destruction flow needs self-destructed target element, service code, data, file system or physical hardware, specially：According to certainly
Instruction or remote self-destruction instruction are ruined, according to preset self-destruction component decision tree, the self-destruction flow of each component and service is calculated, patrols
Position and physical location are collected, and being destroyed according to the self-destruction flow need to self-destructed target element, service code, data, file system
System or physical hardware.
Therefore the self-desttruction equipment and method of a kind of artificial intelligence behavior body provided in an embodiment of the present invention, it can be automatic
The event information of artificial intelligence behavior body is acquired, and judges whether it is in runaway condition, if it needs to execute self-destruction, if
It is that self-destruction is then carried out to each component, service and system by self-destruction module, realizes multi-level, fine-grained artificial intelligence behavior
Body self is destroyed.In addition, self-desttruction equipment can also receive the remote self-destruction instruction of remote equipment transmission, practicability and peace are improved
Quan Xing.
One of ordinary skill in the art will appreciate that realizing all or part of flow in above-described embodiment method, being can be with
Relevant hardware is instructed to complete by computer program, the program can be stored in a computer read/write memory medium
In, the program is when being executed, it may include such as the flow of the embodiment of above-mentioned each method.Wherein, the storage medium can be magnetic
Dish, CD, read-only memory (Read-Only Memory, ROM) or random access memory (Random Access
Memory, RAM) etc..
The above is the preferred embodiment of the present invention, it is noted that for those skilled in the art
For, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications are also considered as
Protection scope of the present invention.
Claims (10)
1. a kind of self-desttruction equipment of artificial intelligence behavior body, which is characterized in that including：Sensing module, decision-making module, self-destruction module
And receiving module；
Wherein, the decision-making module is connect with the self-destruction module, sensing module respectively；The self-destruction module and the reception mould
Block connects；
The sensing module is used to acquire the event information of the artificial intelligence behavior body, and the event information of acquisition is sent to
The decision-making module；
The decision-making module is used for the event information according to the acquisition, judges whether to need to execute self-destruction；If it is, according to
Judging result generates command-destruct, and the command-destruct is sent to self-destruction module；If it is not, then return and continue into
Row judges；
The receiving module is used to receive the remote self-destruction instruction of remote equipment transmission；Wherein, the remote self-destruction instruction has
Highest execution priority；
The self-destruction module is used to be instructed according to the command-destruct or remote self-destruction, calculates each component and the self-destruction stream of service
Journey, logical place and physical location, and being destroyed according to the self-destruction flow need to self-destructed target element, service code, data, text
Part system or physical hardware.
2. the self-desttruction equipment of artificial intelligence behavior body according to claim 1, which is characterized in that the decision-making module is used for root
According to the event information of the acquisition, judge whether to execute self-destruction, specially：
The decision-making module carries out principal component analysis and feature extraction to the event information of the acquisition, obtains the safety after dimensionality reduction
Attribute obtains the result of decision in conjunction with rough set theory, is judged whether to execute self-destruction according to the result of decision.
3. the self-desttruction equipment of artificial intelligence behavior body according to claim 2, which is characterized in that the decision-making module is to described
The event information of acquisition carries out principal component analysis and feature extraction, the security attribute after dimensionality reduction is obtained, in conjunction with rough set theory
Obtain the result of decision, specially：
If the decision condition number of the artificial intelligence behavior body is r, collected security incident sum is N, each safe thing
Part has r dimensions, N number of vector to be expressed as：x1, x2,…,xn, then the safety condition attribute matrix X of n rows r row is constituted；
The decision-making module calculates the average vector and covariance matrix of the safety condition attribute matrix X, then calculates the association
The characteristic value and feature vector of variance matrix are ranked up according to the characteristic value is descending, m pivot before calculated for rank
Contribution rate of accumulative total；The contribution rate of accumulative total is for the component after weighing dimensionality reduction to the fidelity of initial data；
The feature vector corresponding to the preceding m pivot is extracted, constitutes transformation matrix T, and the preceding m are calculated by Y=XT
The principal component of pivot obtains the result of decision in conjunction with rough set theory.
4. the self-desttruction equipment of artificial intelligence behavior body according to claim 1, which is characterized in that the self-destruction module includes
Self-destruction executive module and self-destruction control assembly；
The self-destruction executive module is used to be instructed according to the command-destruct or remote self-destruction, according to preset self-destruction component decision
Tree calculates each component and self-destruction flow, logical place and the physical location of service, and being destroyed according to the self-destruction flow need to be certainly
Target element, service code, data, file system or the physical hardware ruined；
The self-destruction control assembly is used to confirm the command-destruct or the credibility of remote self-destruction instruction, and monitors entire self-destruction
Flow is intervened when self-destruction is failed.
5. the self-desttruction equipment of artificial intelligence behavior body according to claim 1, which is characterized in that the self-desttruction equipment concatenation
Between power supply system and artificial intelligence decision system.
6. a kind of self-destruction method of artificial intelligence behavior body, which is characterized in that including：
Acquire the event information of the artificial intelligence behavior body；
According to the event information of the acquisition, judge whether to need to execute self-destruction；If it is, according to judging result, generate certainly
Instruction is ruined, and the command-destruct is sent to self-destruction module；If it is not, then returning and continuing to judge；
According to the command-destruct, each component and self-destruction flow, logical place and the physical location of service are calculated, and according to institute
Stating the destruction of self-destruction flow needs self-destructed target element, service code, data, file system or physical hardware.
7. the self-destruction method of artificial intelligence behavior body according to claim 6, which is characterized in that the self-destruction method also wraps
It includes：
Receive the remote self-destruction instruction that remote equipment is sent；
It is instructed according to the remote self-destruction, judges whether the remote self-destruction instruction is credible, if remote self-destruction instruction comes
Source is credible, then calculates each component and self-destruction flow, logical place and the physical location of service, and according to the self-destruction flow pin
Ruining needs self-destructed target element, service code, data, file system or physical hardware.
8. the self-destruction method of artificial intelligence behavior body according to claim 7, which is characterized in that described according to the acquisition
Event information, judge whether to need to execute self-destruction, specially：Principal component analysis and spy are carried out to the event information of the acquisition
Sign extraction, obtains the security attribute after dimensionality reduction, the result of decision is obtained in conjunction with rough set theory, is judged according to the result of decision
Whether self-destruction is executed；
Wherein, the event information of the acquisition includes the status information of artificial intelligence behavior body itself.
9. the self-destruction method of artificial intelligence behavior body according to claim 8, which is characterized in that described to the acquisition
Event information carries out principal component analysis and feature extraction, obtains the security attribute after dimensionality reduction, is obtained certainly in conjunction with rough set theory
Plan is as a result, be specially：
If the decision condition number of the artificial intelligence behavior body is r, collected security incident sum is N, each safe thing
Part has r dimensions, N number of vector to be expressed as：x1, x2,…,xn, then the safety condition attribute matrix X of n rows r row is constituted；
The average vector and covariance matrix of the safety condition attribute matrix X are calculated, then calculates the spy of the covariance matrix
Value indicative and feature vector are ranked up according to the characteristic value is descending, the contribution rate of accumulative total of m pivot before calculated for rank；
The contribution rate of accumulative total is for the component after weighing dimensionality reduction to the fidelity of initial data；
The feature vector corresponding to the preceding m pivot is extracted, constitutes transformation matrix T, and the preceding m are calculated by Y=XT
The principal component of pivot obtains the result of decision in conjunction with rough set theory.
10. the self-destruction method of artificial intelligence behavior body according to claim 6, which is characterized in that described to calculate each group
Self-destruction flow, logical place and the physical location of part and service, and destroy according to the self-destruction flow need self-destructed target element,
Service code, data, file system or physical hardware, specially：
It calculates the self-destruction flow of each component and service according to preset self-destruction component decision tree according to the command-destruct, patrol
Position and physical location are collected, and being destroyed according to the self-destruction flow need to self-destructed target element, service code, data, file system
System or physical hardware.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810234182.7A CN108460452A (en) | 2018-03-20 | 2018-03-20 | A kind of self-desttruction equipment and method of artificial intelligence behavior body |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810234182.7A CN108460452A (en) | 2018-03-20 | 2018-03-20 | A kind of self-desttruction equipment and method of artificial intelligence behavior body |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108460452A true CN108460452A (en) | 2018-08-28 |
Family
ID=63237385
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810234182.7A Pending CN108460452A (en) | 2018-03-20 | 2018-03-20 | A kind of self-desttruction equipment and method of artificial intelligence behavior body |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108460452A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112434901A (en) * | 2020-10-15 | 2021-03-02 | 合肥工业大学 | Intelligent re-decision method and system for traffic patrol scheme of unmanned aerial vehicle |
| CN113467440A (en) * | 2021-06-09 | 2021-10-01 | 珠海云洲智能科技股份有限公司 | Navigation controller |
| CN116756785A (en) * | 2023-08-16 | 2023-09-15 | 中国兵器装备集团兵器装备研究所 | Self-destruction method and system of intelligent terminal system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101477604A (en) * | 2009-01-16 | 2009-07-08 | 哈尔滨工程大学 | Layered self-destruction method oriented to distributed system |
| CN103234255A (en) * | 2013-04-16 | 2013-08-07 | 杭州汉超科技有限公司 | Remote positioning, monitoring and alarming cabinet air conditioner with self-destruction device and control method thereof |
| US20140116234A1 (en) * | 2012-08-23 | 2014-05-01 | Emmanuel Daniel Martin Jacq | Autonomous unmanned tower military mobile intermodal container and method of using the same |
| CN106778370A (en) * | 2016-12-23 | 2017-05-31 | 四川九洲电器集团有限责任公司 | The self-destruction method and equipment of a kind of equipment for being capable of autonomous |
-
2018
- 2018-03-20 CN CN201810234182.7A patent/CN108460452A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101477604A (en) * | 2009-01-16 | 2009-07-08 | 哈尔滨工程大学 | Layered self-destruction method oriented to distributed system |
| US20140116234A1 (en) * | 2012-08-23 | 2014-05-01 | Emmanuel Daniel Martin Jacq | Autonomous unmanned tower military mobile intermodal container and method of using the same |
| CN103234255A (en) * | 2013-04-16 | 2013-08-07 | 杭州汉超科技有限公司 | Remote positioning, monitoring and alarming cabinet air conditioner with self-destruction device and control method thereof |
| CN106778370A (en) * | 2016-12-23 | 2017-05-31 | 四川九洲电器集团有限责任公司 | The self-destruction method and equipment of a kind of equipment for being capable of autonomous |
Non-Patent Citations (4)
| Title |
|---|
| KARTHIKEYAN M ET AL.: "Intelligent Exploration and Surveillance Robot In Defense Environment", 《INTERNATIONAL JOURNAL OF ADVANCED RESEARCH IN ELECTRICAL, ELECTRONICS AND INSTRUMENTATION ENGINEERING》 * |
| 张延松 等: "基于PCA和粗糙集构建决策树的变电站故障诊断", 《电力系统保护与控制》 * |
| 王国胤 等: "粗糙集理论与应用研究综述", 《计算机学报》 * |
| 邱志涛 等: "分布式系统层次化自毁方法及案例研究", 《计算机应用研究》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112434901A (en) * | 2020-10-15 | 2021-03-02 | 合肥工业大学 | Intelligent re-decision method and system for traffic patrol scheme of unmanned aerial vehicle |
| CN112434901B (en) * | 2020-10-15 | 2022-11-04 | 合肥工业大学 | Intelligent re-decision method and system for traffic patrol scheme of unmanned aerial vehicle |
| CN113467440A (en) * | 2021-06-09 | 2021-10-01 | 珠海云洲智能科技股份有限公司 | Navigation controller |
| CN116756785A (en) * | 2023-08-16 | 2023-09-15 | 中国兵器装备集团兵器装备研究所 | Self-destruction method and system of intelligent terminal system |
| CN116756785B (en) * | 2023-08-16 | 2023-11-10 | 中国兵器装备集团兵器装备研究所 | Self-destruction method and system of intelligent terminal system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Singh | Learning based driver drowsiness detection model | |
| CN108460452A (en) | A kind of self-desttruction equipment and method of artificial intelligence behavior body | |
| US10677563B2 (en) | Firearm usage monitoring system | |
| US20200003511A1 (en) | Firearm usage monitoring system | |
| US11420655B2 (en) | System and method for safety and efficacy override of an autonomous system | |
| US20200011629A1 (en) | Firearm usage monitoring system | |
| Schulzke | Autonomous weapons and distributed responsibility | |
| US11408699B2 (en) | Firearm usage monitoring system | |
| Liu | The power structure of artificial intelligence | |
| CN103839379A (en) | Automobile and driver fatigue early warning detecting method and system for automobile | |
| DE102014225332A1 (en) | System, in particular hand-held power tool system | |
| EP3574278A1 (en) | Firearm usage monitoring system | |
| Cass | Autonomous weapons and accountability: Seeking solutions in the law of war | |
| CN109559008A (en) | Construction monitoring method, apparatus and system | |
| Abraham et al. | Adaptive autonomy in human-on-the-loop vision-based robotics systems | |
| Raman et al. | Towards minimal explanations of unsynthesizability for high-level robot behaviors | |
| Johnson | The AI commander problem: Ethical, political, and psychological dilemmas of human-machine interactions in AI-enabled warfare | |
| CN113778113B (en) | Pilot auxiliary driving method and pilot auxiliary driving system based on multi-mode physiological signals | |
| Lattner et al. | Knowledge-based risk assessment for intelligent vehicles | |
| Li et al. | Model calibration concerning risk coefficients of driving safety field model | |
| Koncz et al. | Fuzzy rule-based hierarchical overall risk analysis of battery testing laboratories | |
| Duarte et al. | Robot Self-defense: Robot, don't hurt me, no more | |
| Brown | Making sense of the noise: an ABC approach to big data and security | |
| CN110126723B (en) | Lock control system, method and device, storage medium and electronic device | |
| CN113806828A (en) | Cover opening protection device and method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180828 |