KR20220029434A - System and Method for Selecting Compliance Data for Complicance Modeling of a UAV - Google Patents

Abstract

The present invention relates to a system and method for selecting compliance data for building a compliance model to detect whether there is a normal action of an unmanned aerial vehicle. The system for selecting compliance data for compliance modeling of an unmanned aerial vehicle according to the present invention comprises: an action rule defining unit configured to define action rules the unmanned aerial vehicle is allowed; a compliance data candidate defining unit configured to define compliance data candidate parameters; an action rule violation indicator defining unit configured to define a violation indicator for each of the action rules; a candidate parameter deriving unit configured to derive the compliance data candidate parameters corresponding to the action rule violation indicator and a threshold thereof; a candidate parameter data collecting unit configured to collect compliance history data representing compliance states of the candidate parameters every set period; a compliance instance calculating unit configured to calculate a compliance instance as a ratio of a compliance period to the entire collection period based on the compliance history data of the candidate parameters; and a fitness calculating unit configured to select or not to select the candidate parameters as compliance data by calculating the goodness of fit of the candidate parameters based on the compliance instance.

Description

{System and Method for Selecting Compliance Data for Compliance Modeling of a UAV}

The present invention relates to compliance modeling of an unmanned aerial vehicle, and more particularly, to a system and method for selecting compliance data for compliance modeling for detecting whether an unmanned aerial vehicle behaves normally.

Unmanned Aerial Vehicle (UAV) is a generic term for military drones in the shape of airplanes or helicopters that can be flown and controlled by induction of radio waves without a pilot.

1 is a block diagram illustrating a data transmission system between a general ground control system and an unmanned aerial vehicle.

The data transmission system includes an up link for wirelessly transmitting command or control information from the ground system 110 to the unmanned aerial vehicle 120 , and the unmanned aerial vehicle 120 to the ground system 110 . It is made possible to transmit and receive wireless data through a down link that wirelessly transmits status information of the unmanned aerial vehicle 120 and the mounted device.

The ground system 110 generates control information to be transmitted to the unmanned aerial vehicle 120 and includes a ground control unit 111 that encrypts the generated control information, and modulates the encrypted control information to create the unmanned aerial vehicle 120 . and a ground communication device 112 that transmits to the aircraft 120 .

The unmanned aerial vehicle 120 generates state information of the unmanned aerial vehicle 120 to be transmitted to the ground system 110 and includes a flight controller 122 that encrypts the generated control information; It is configured to include a flight communication device 121 that modulates the state information and transmits it to the ground system 110 .

The ground communication device 112 and the flight communication device 121 may modulate and transmit an encrypted signal, respectively, and receive the encrypted and modulated signal and demodulate it into a desired signal.

In general, an unmanned aerial vehicle sets a flight path for performing a mission given from a ground system, moves along the flight path, performs a mission, and then returns to an original position. However, due to factors such as wind, battery condition, temperature, propeller speed, weight, and the like while performing a mission, it is very difficult for an unmanned aerial vehicle to accurately follow a set flight path. Therefore, even while the unmanned aerial vehicle normally performs a mission, a certain distance error and behavioral error occur between the set flight path and the actual flight path of the unmanned aerial vehicle, and between the setting action and the actual action.

On the other hand, unmanned aerial vehicles are rapidly growing in their utility in the private sector, such as broadcast filming and delivery of goods, and are a technology that has a very high utility in military applications such as surveillance/reconnaissance and detection. There is a vulnerability in wireless security for wireless communication between the unmanned aerial vehicle and the ground system. In other words, hacking that disrupts or sniffs the wireless signal between the unmanned aerial vehicle and the ground system, forcibly cuts off communication between the ground system and the unmanned aerial vehicle, connects to the unmanned aerial vehicle with unauthorized equipment, and arbitrarily controls the operation of the unmanned aerial vehicle. can occur

If the unmanned aerial vehicle is abused due to hacking of the unmanned aerial vehicle, it may pose a serious threat to national security, such as delivery of goods being delivered to another location or self-destructive attack on important national facilities. In order to prevent this, conventionally, various anti-hacking technologies have been developed, or various anti-drone technologies have been proposed to neutralize the malicious operation of an exploited unmanned aerial vehicle.

When malicious hacking is applied to the unmanned aerial vehicle and the unmanned aerial vehicle deviates from the normal flight path and flies to another path, or performs malicious behavior outside the set behavior, it is necessary to detect and detect it in advance.

However, as described above, even while the unmanned aerial vehicle is performing its mission normally, a distance error occurs between the set flight path and the actual flight path, and a behavioral error occurs between the setting action and the actual action. Even when flying, a distance error occurs between the set flight path and the actual flight path, and a behavioral error occurs between the set action and the actual action. For this reason, it is very difficult to accurately determine whether the current behavior of the unmanned aerial vehicle is normal based on the location information or posture information of the unmanned aerial vehicle. A technology that analyzes whether an act is malicious in real time has not been proposed.

The applicant of the present invention proposes a technology for detecting malicious behavior of an unmanned aerial vehicle in advance through compliance modeling indicating how well the current actual flight path and behavior of the unmanned aerial vehicle follow the set flight path and set behavior.

2 is a diagram illustrating a system for analyzing compliance of a general unmanned aerial vehicle. The conformance analysis system of the unmanned aerial vehicle may be implemented with the conformance model module 21 having completed learning the conformance data of the normal behavior and the conformance data of the malicious act. The compliance model module 21 learns whether the corresponding compliance data is a normal behavior or a malicious behavior with respect to a plurality of compliance data collected in various environments.

If the compliance data newly collected from the unmanned aerial vehicle is applied to the learned compliance model module 21, it is possible to determine in real time whether the corresponding unmanned aerial vehicle is acting normally or maliciously.

The applicant of the present invention has defined the behavior rules and security characteristics applicable to the military unmanned aerial vehicle as shown in [Table 1] below. These rules of conduct may be defined in various ways according to the function and role of the unmanned aerial vehicle.

[Table 1]

It is necessary to collect and learn different compliance data for each of the above-described behavior rules to determine whether or not to comply with each behavior rule.

Various sensor modules are installed in the unmanned aerial vehicle, and the data that can be collected by each sensor can be organized as shown in [Table 2] below.

[Table 2]

The behavior rules of the unmanned aerial vehicle can be defined in many ways, and the data collected from the unmanned aerial vehicle are also very diverse. Compliance modeling should be done based on appropriate compliance data for each of the various behavior rules. The first action rule BR 1 is 'Need to sail on a designated correct route', and the data for determining whether the unmanned aerial vehicle complies with the first action rule is location information. The current sensing data is unnecessary to determine whether the first behavior rule is complied with. On the other hand, current sensing data is required to determine whether the unmanned aerial vehicle complies with the 11th behavior rule (BR 11) 'consuming minimal energy when loitering'.

As described above, compliance modeling for each of the various behavior rules should be done separately, and optimal compliance data should be selected for each compliance modeling.

Republic of Korea Patent Registration No. 10-2194734 Republic of Korea Patent Publication No. 2017-0074369

The present invention has been devised to solve the above needs, and it is an object of the present invention to provide a system and method for selecting compliance data suitable for compliance modeling for determining whether an unmanned aerial vehicle is performing an action to comply with a set action.

However, the technical task to be achieved by the present embodiment is not limited to the technical task as described above, and other technical tasks may exist.

A system for selecting compliance data for conformance modeling of an unmanned aerial vehicle according to an embodiment of the present invention comprises: an action rule definition unit defining an action rule allowed for an unmanned aerial vehicle;

a conformance data candidate definition unit defining conformance data candidate parameters;

a behavior rule violation indicator definition unit defining a violation indicator for each behavior rule;

a candidate parameter derivation unit for deriving the compliance data candidate parameter corresponding to the behavior rule violation indicator and a threshold value thereof;

a candidate parameter data collection unit that collects compliance history data indicating the compliance status of the candidate parameter at regular intervals;

a compliance instance calculator for calculating compliance instances as a ratio of the compliance period among the entire collection period based on the compliance history data of the candidate parameters;

and a fitness calculator configured to calculate the fitness of the candidate parameter based on the compliance instance and select or non-select the candidate parameter as compliance data.

In more detail, the compliance data candidate parameter is characterized in that it is defined based on sensors mounted on the unmanned aerial vehicle.

In more detail, the candidate parameter derivation unit may further define two or more candidate parameters and a logical expression between the two or more candidate parameters.

More specifically, the compliance instance is characterized in that the ratio of the number of times of compliance to the total number of collections.

More specifically, the fitness calculator uses a random variable X having a distribution of G(X)=Beta(α,β) to model the fitness, and α,β are the better distribution parameters, and the best likelihood method for the compliance instance It is characterized in that it is estimated by applying

In addition, the method for selecting compliance data for conformance modeling of an unmanned aerial vehicle according to an embodiment of the present invention selects stored compliance data in which behavior rules allowed for the unmanned aerial vehicle, compliance data candidate parameters, and violation indicators for each behavior rule are defined A method for selecting compliance data in a system, the method comprising:

A first step of deriving, by a candidate parameter derivation unit, the compliance data candidate parameter corresponding to the violation index for each behavior rule and a threshold value thereof;

a second step of collecting, by a candidate parameter data collection unit, compliance history data indicating the compliance status of the candidate parameter at regular intervals;

A third step in which the compliance instance calculation unit calculates the compliance instances as a percentage of the compliance period among the entire collection period based on the compliance history data of the candidate parameters, and

and a fourth step in which the fitness calculator calculates the fitness of the candidate parameter based on the compliance instance and selects or deselects the candidate parameter as compliance data.

According to the present invention, by selecting optimal compliance data for each behavior rule conformance modeling of the unmanned aerial vehicle, the degree of compliance for each behavior rule is accurately determined to determine whether the unmanned aerial vehicle performs a normal behavior conforming to the behavior rule or deviates from the behavior rule. There is an advantage in being able to accurately determine whether an act of evil is being carried out.

1 is a block diagram illustrating a data transmission system between a general ground control system and an unmanned aerial vehicle.
2 is a diagram illustrating a system for analyzing compliance of a general unmanned aerial vehicle.
3 is a block diagram illustrating a modeling data selection system for modeling compliance of an unmanned aerial vehicle according to an embodiment of the present invention.
4 is an operation flowchart illustrating a method for selecting compliance data for compliance modeling of an unmanned aerial vehicle according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those of ordinary skill in the art can easily implement them. However, the present invention may be embodied in various different forms and is not limited to the embodiments described herein, and it should be understood to include all modifications, equivalents, and substitutes included in the spirit and scope of the present invention. In order to clearly explain the present invention, parts irrelevant to the description in the drawings are omitted, and similar reference numerals are attached to similar parts throughout the specification. In addition, while describing with reference to the drawings, even if the configuration indicated by the same name, the drawing number may vary depending on the drawing, and the drawing number is only described for convenience of description, and the concept, feature, function of each configuration by the corresponding drawing number Or the effect is not to be construed as limiting.

Throughout the specification, when a part is "connected" with another part, this includes not only the case of being "directly connected" but also the case of being "electrically connected" with another element interposed therebetween. . In addition, when a part "includes" a certain component, it means that other components may be further included, rather than excluding other components, unless specifically stated to the contrary. It should be understood that this does not preclude the possibility of addition or presence of features or numbers, steps, operations, components, parts, or combinations thereof.

As used herein, the term 'part' or 'module' includes a unit realized by hardware or software, and a unit realized using both, and one unit is realized using two or more hardware Alternatively, two or more units may be implemented by one piece of hardware.

Unless defined otherwise, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Terms such as those defined in commonly used dictionaries should be interpreted as having meanings consistent with the meanings in the context of the related art, and should not be interpreted in an ideal or excessively formal meaning unless explicitly defined in the present application. shouldn't

3 is a block diagram illustrating a system for selecting compliance data for modeling compliance of an unmanned aerial vehicle according to an embodiment of the present invention.

The compliance data selection system of the present invention includes an action rule definition unit 31 defining an action rule allowed for an unmanned aerial vehicle, and a compliance data candidate definition unit defining a compliance data candidate parameter based on sensors mounted on the unmanned aerial vehicle. (32), a behavior rule violation indicator defining unit 33 defining a violation indicator for each behavior rule, and a candidate parameter derivation unit for deriving the compliance data candidate parameter corresponding to the behavior rule violation indicator and a threshold value thereof ( 34), a candidate parameter data collection unit 35 that collects compliance history data indicating the compliance status of the candidate parameters at regular intervals, and based on the compliance history data of the candidate parameters as a percentage of the compliance period among the entire collection period and a conformance instance calculator 36 for calculating conformance instances, and a conformance calculator 37 for selecting or not selecting the candidate parameter as conformance data by calculating the fit of the candidate parameter based on the conformance instance.

The behavior rule definition unit 31 defines the behavior rules of the unmanned aerial vehicle as shown in [Table 1] above. [Table 1] shows the behavior rules defined by the applicant of the present invention for the military unmanned aerial vehicle, and the behavior rules of the unmanned aerial vehicle defined in the behavior rule definition unit 31 may be modified according to the role and function of the corresponding unmanned aerial vehicle.

The compliance data candidate definition unit 32 defines the compliance data candidate parameters based on sensors mounted on the unmanned aerial vehicle. The unmanned aerial vehicle is equipped with a plurality of sensors as shown in [Table 2], and based on this, candidate parameters for compliance data can be defined as shown in [Table 3] below.

[Table 3]

The rule violation indicator definition unit 33 defines an indicator that violates the corresponding behavior rule for each behavior rule defined in the behavior rule definition unit 31 . For example, the first action rule (BR1) is 'Need to sail on a designated correct route', and the indicator that violates this first action rule is '|Location-Planned Location| > D _distance '. That is, if the distance difference between the current location and the designated location is out of the threshold, this corresponds to a violation of the first action rule. Similarly, for each behavior rule, an indicator that violates the corresponding behavior rule is defined.

The candidate parameter derivation unit 34 derives a compliance data candidate parameter and a threshold value corresponding to the behavior rule violation index. When it is necessary to compare the threshold distance and the difference between the current location and the designated route location from the first behavior rule violation indicator, latitude, longitude, altitude, etc. may be included as candidate parameters for compliance data, and a threshold value is derived for each candidate parameter can be This threshold value may be an absolute value or may be a threshold difference value from latitude, longitude, and altitude of a set route. When two or more candidate parameters are derived, their logical expressions are also defined. That is, whether or not to OR or multiply the result of comparing the latitude difference value and the threshold value and the longitude difference value and the threshold value can be additionally derived.

The candidate parameter data collection unit 35 collects compliance history data indicating the compliance status of the corresponding candidate parameter at regular intervals. That is, if the data of the candidate parameter is within the threshold value, the compliance history data becomes '1', because the data of the candidate parameter is within the threshold value.

Since the candidate parameter data collection unit 35 does not collect the data values of the candidate parameters in this way, but collects compliance history data indicating whether the data of the corresponding candidate parameters conforms to the behavior rule indicator, the collected data can be reduced.

The compliance instance calculator 36 calculates the compliance instance as a ratio of the compliance period among the entire collection period based on the compliance history data of the candidate parameters. The compliance instance calculator 36 may calculate the compliance instance as a ratio of the number of times of compliance to the total number of collections.

The fitness calculator 37 calculates the fitness of the candidate parameter as compliance data based on the compliance instance, and selects or non-selects the candidate parameter as the compliance data. When there are many candidate parameters previously selected, a candidate parameter optimal for compliance modeling may be selected as compliance data through this fitness calculation process.

The fitness calculator 37 may model the fitness using a random variable X having a distribution of G(X)=Beta(α,β). Here, α and β are better distribution parameters, which can be estimated by applying the most likelihood method to the compliance instance. The expected value of the random variable X is calculated as (α/(α+β)). If this value is greater than the minimum fit threshold, C _T , the corresponding candidate parameter is selected as compliance data. If not, the corresponding candidate parameter is not selected as compliance data.

When the compliance data is selected in this way, as shown in FIG. 2, the compliance data of the normal behavior state and the compliance data of the malicious behavior are collected respectively to learn the compliance model module 21, and the learned compliance model module 21 By applying the newly collected compliance data, it is possible to determine in real time whether the unmanned aerial vehicle is behaving normally or maliciously.

The compliance data selection system of the present invention may be implemented in any computing system. For example, this compliance data selection system may be implemented in a stand alone computing system, or may be implemented in distributed computing systems capable of communicating with each other via a network or the like. In addition, the conformance data selection system may include a part implemented by a processor executing a program including a series of instructions, or a part implemented by logic hardware designed by logic synthesis. . In the present specification, a processor is hardware-implemented, including circuits physically structured to execute predefined operations including operations expressed in code and/or instructions included in a program. It may refer to any data processing device. For example, the data processing apparatus includes a microprocessor, a central processing unit (CPU), a graphics processing unit (GPU), a neural processing unit (NPU), a processor core, a multi-core processor, a multi-processor, and an application-specific integrated (ASIC). circuit), an application-specific instruction-set processor (ASIP), and a field programmable gate array (FPGA).

4 is an operation flowchart illustrating a method for selecting compliance data for compliance modeling of an unmanned aerial vehicle according to an embodiment of the present invention.

Behavior rules allowed in the unmanned aerial vehicle, compliance data candidate parameters based on sensors installed in the unmanned aerial vehicle, and violation indicators for each behavior rule are defined and stored (S41).

The candidate parameter derivation unit derives the threshold value of the compliance candidate parameter corresponding to the violation index for each behavior rule (S42).

The candidate parameter data collection unit 35 collects compliance history data indicating the compliance status of the candidate parameter at regular intervals (S43).

The compliance instance calculator 36 calculates the compliance instance as a ratio of the compliance period among the entire collection period based on the compliance history data of the candidate parameter ( S44 ).

The fitness calculator 37 calculates the fitness of the candidate parameter based on the compliance instance and selects or deselects the candidate parameter as compliance data (S45).

The description of the present invention described above is for illustration, and those of ordinary skill in the art to which the present invention pertains can understand that it can be easily modified into other specific forms without changing the technical spirit or essential features of the present invention. will be. Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive. For example, each component described as a single type may be implemented in a dispersed form, and likewise components described as distributed may also be implemented in a combined form.

The scope of the present invention is indicated by the following claims rather than the above detailed description, and all changes or modifications derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present invention. do.

31: behavior rule definition unit 32: conformance data candidate definition unit
33: behavior rule violation indicator definition unit 34: candidate parameter derivation unit
35: candidate parameter data collection unit 36: compliance instance calculation unit
37: fitness calculator

Claims (10)

a behavior rule definition unit defining a behavior rule allowed for the unmanned aerial vehicle;
a conformance data candidate definition unit defining conformance data candidate parameters;
a behavior rule violation indicator definition unit defining a violation indicator for each behavior rule;
a candidate parameter derivation unit for deriving the compliance data candidate parameter corresponding to the behavior rule violation indicator and a threshold value thereof;
a candidate parameter data collection unit that collects compliance history data indicating the compliance status of the candidate parameter at regular intervals;
a compliance instance calculator for calculating compliance instances as a ratio of the compliance period among the entire collection period based on the compliance history data of the candidate parameters;
and a fitness calculator configured to calculate the fitness of the candidate parameter based on the compliance instance and select or not select the candidate parameter as compliance data. The system of claim 1 , wherein the compliance data candidate parameter is defined based on sensors mounted on the unmanned aerial vehicle. The system of claim 1 , wherein the candidate parameter derivation unit further defines two or more candidate parameters and a logical expression between the two or more candidate parameters. The system of claim 1 , wherein the compliance instance is a ratio of the number of times of compliance to the total number of collections. The method according to claim 1, wherein the fitness calculator models the fitness using a random variable X having a distribution of G(X)=Beta(α,β), wherein α,β are a beta distribution parameter, and the best likelihood method for compliance instances Compliance data selection system for compliance modeling of an unmanned aerial vehicle, characterized in that it is estimated by applying A method for selecting compliance data in a compliance data selection system in which an action rule allowed for an unmanned aerial vehicle, a compliance data candidate parameter, and a violation index for each action rule are defined and stored,
A first step of deriving, by a candidate parameter derivation unit, the compliance data candidate parameter corresponding to the violation index for each behavior rule and a threshold value thereof;
a second step of collecting, by a candidate parameter data collection unit, compliance history data indicating the compliance status of the candidate parameter at regular intervals;
A third step in which the compliance instance calculation unit calculates the compliance instances as a percentage of the compliance period among the entire collection period based on the compliance history data of the candidate parameters, and
and a fourth step of selecting or not selecting the candidate parameter as compliance data by a fitness calculator calculating the fitness of the candidate parameter based on the compliance instance. The method of claim 6 , wherein the compliance data candidate parameter is defined based on sensors mounted on the unmanned aerial vehicle. The method of claim 6, wherein the candidate parameter derivation unit further defines two or more candidate parameters and a logical expression between the two or more candidate parameters. The method of claim 6 , wherein the compliance instance is a ratio of the number of times of compliance to the total number of collections. 7. The method according to claim 6, wherein the fitness calculator uses a random variable X having a distribution of G(X)=Beta(α,β) to model the goodness of fit, wherein α,β are parameters of a better distribution, and the best likelihood method for compliance instances A method of selecting compliance data for compliance modeling of an unmanned aerial vehicle, characterized in that it is estimated by applying

Images