KR102614755B1 - Optimal UAV Movement Path Planning and Sensing Data Acquisition Method Using PSO-based Machine Learning in UAV Assisted Sensor Data Acquisition Environment - Google Patents

Abstract

A method and system for obtaining optimal movement path and sensing information using PSO-based biomimetic machine learning is presented. The system proposed in the present invention uses information on pre-stored sensor deployment maps, 3D terrain maps, and 3D no-fly zone maps to calculate the evaluation value of the entire objective function for the candidates for the UAV's next sensing information acquisition location. A UAV coverage calculation unit that calculates the range of sensor area grids that can communicate at a location next to the UAV, a sensing value objective function calculation unit that calculates the value of sensing information obtainable from sensors within the UAV coverage at the calculated location next to the UAV, UAV operating time objective function calculation unit, which calculates the objective function value by estimating the flight time required to move from the current location of the UAV to the next location and the time required to acquire sensing information at the next location, from the current location of the UAV Constraints required for UAV network operation using the UAV operation energy consumption objective function calculation unit and PSO algorithm that calculates the objective function value by estimating the energy required to move to the next location and the energy required to acquire sensing information at the next location. It includes a PSO-based overall objective function maximum value UAV position calculation unit that satisfies the conditions and derives the position of the UAV that can maximize the calculated overall objective function.

Description

Optimal UAV Movement Path Planning and Sensing Data Acquisition Method Using PSO-based Machine Learning in UAV Assisted Sensor Data Acquisition Environment}

The present invention relates to a method and system for acquiring optimal movement path and sensing information using PSO-based biomimetic machine learning in sensor data acquisition using an unmanned aircraft.

A wireless sensor is a device that monitors the surrounding environment using various sensors and can detect or track specific events. It is a technology that can transmit the acquired sensing information to another device using a wireless transmission device.

In an outdoor environment, a sensor network or existing communication infrastructure can be used as a method to transmit information from sensors placed in a three-dimensional space to a remote server. However, it is difficult to guarantee connectivity with only sensor devices, or existing communication infrastructure cannot be used. If there is no such technology, a technology that uses an unmanned aircraft to receive information from sensors in real time and configures a network between unmanned aircraft to transmit all data to a remote server can be used.

In order to acquire sensing information obtained from sensors on the ground using multiple unmanned aerial vehicles, the unmanned aerial vehicles must be located at a height where the unmanned aerial vehicles and sensors on the ground can communicate, and the range of ground area that can communicate according to height and the quality of communication with the sensors must be determined. This changes. Additionally, when an unmanned aircraft moves at a high speed, it is difficult to obtain sufficient sensing information smoothly from sensors on the ground, so it is necessary to stop at a specific location, obtain sensing information from sensors within the communication range, and then move to the next location.

The value of information obtained from sensors varies depending on what type of sensor it is and how much time has passed since the previous sensing information was acquired. Therefore, when positioning unmanned aerial vehicles, it is necessary to consider where and at what height the highest sensing value is obtained. Since unmanned aerial vehicles must acquire a lot of information within a limited time, information acquisition time and energy consumption are very important factors. Accordingly, a method is needed to maximize the value of information and minimize time and energy consumption.

When operating multiple unmanned aerial vehicles, asynchronous optimization is necessary because the unmanned aerial vehicles move asynchronously, and the location of each aerial vehicle must be determined considering the sensing information and area information obtained by other unmanned aerial vehicles. In addition, sensing information is transmitted to a remote server through an unmanned aerial vehicle network, and the remote server processes the sensing information and determines the next optimal location for the unmanned aerial vehicles and transmits it to the unmanned aerial vehicles. To this end, connectivity between unmanned aircraft must be guaranteed in a given environment.

Korean Patent No. 10-2009608 (2019.08.05)

The technical problem to be achieved by the present invention is to acquire sensing information from sensors placed in a three-dimensional outdoor space in real time using a plurality of unmanned aerial vehicles (UAVs), and to determine the value of the sensing information that can be obtained from UAVs. The goal is to provide a method and system for deriving optimal 3D location information and movement paths that can maximize UAV operation time and UAV usage energy while simultaneously minimizing it.

In one aspect, the optimal movement path and sensing information acquisition system using PSO-based biomimetic machine learning in sensor data acquisition using an unmanned flying vehicle proposed in the present invention includes a pre-stored sensor placement map, a 3D terrain map, and a 3D flight ban. A UAV coverage calculation unit that calculates the range of sensor area grids that can communicate at the next location of the UAV in order to calculate the evaluation value of the overall objective function for the UAV's next sensing information acquisition location candidates using information about the area map. , a sensing value objective function calculation unit that calculates the value of sensing information that can be obtained from sensors within the UAV coverage at the calculated next location of the UAV, the flight time required to move from the current location of the UAV to the next location, and the next location. A UAV operating time objective function calculation unit that calculates the objective function value by estimating the time required to acquire sensing information, the energy required to move from the current location of the UAV to the next location, and the energy required to acquire sensing information at the next location. UAV operation energy consumption objective function calculation unit that calculates the objective function value by estimating energy, calculates the change in sensing information value according to sensor type for each sensing grid over time and provides the values required for calculation of the sensing information objective function calculation unit A sensing value calculation unit for each sensing grid, a multiple UAV current location data storage unit that stores the current location information of all UAVs placed in the sensor field and provides it to the sensing information objective function calculation unit, and a PSO (Particle Swam Optimization) algorithm. It includes a PSO-based maximum overall objective function UAV position calculation unit that satisfies the constraints required for UAV network operation and derives the position of the UAV that can maximize the calculated overall objective function.

The sensing value objective function calculation unit defines the sensing information value obtainable from all grid cells within the coverage of the UAV as a normalized sensing value objective function, and the UAV operation time objective function calculation unit calculates the operation time from the current location to the next location as follows. The sum of the estimated travel time to move to the location and the estimated time to acquire sensing information from sensors within coverage while staying at the next location is defined as a normalized UAV operation time objective function, and the UAV operation energy consumption objective function calculation unit is currently Operational energy consumption for the next evaluation position is consumed as UAV movement flight energy to move to the next location, UAV stop flight energy for the UAV to fly stationary to acquire sensing information at the next location, and acquisition of sensing information at the next location. For this purpose, the sum of UAV communication energy for communication between UAV and sensors is defined as a normalized UAV operation energy consumption objective function.

The PSO-based overall objective function maximum value UAV position calculation unit performs a search by applying the PSO algorithm to reduce the amount of calculation of the objective function and enable real-time implementation, and after completing the acquisition of sensing data at the current location according to the movement of each UAV, When moving to the next optimal location, the new moving location is selected as the optimal location by classifying positions that satisfy predetermined conditions.

In another aspect, the method of acquiring optimal movement path and sensing information using PSO-based biomimetic machine learning in sensor data acquisition using an unmanned aircraft proposed in the present invention includes a sensor arrangement map in which a UAV coverage calculation unit is pre-stored, In order to calculate the evaluation value of the overall objective function for the candidates for the UAV's next sensing information acquisition location using information on the 3D terrain map and 3D no-fly zone map, the range of the sensor area grids that can communicate at the next location of the UAV is determined. Calculating step, the sensing value objective function calculation section calculates the value of sensing information that can be obtained from sensors within the UAV coverage at the location next to the calculated UAV, the UAV operating time objective function calculation section moves from the current location of the UAV to the next location. A step of calculating the objective function value by estimating the flight time required for movement of the UAV and the time required to acquire sensing information at the next location. The UAV operation energy consumption objective function calculation unit calculates the necessary for movement from the current location of the UAV to the next location. A step of calculating the objective function value by estimating the energy and the energy required to acquire sensing information at the next location. The sensing value calculation unit for each sensing grid calculates the change in sensing information value according to the sensor type for each sensing grid over time. A step of providing the values necessary for calculation of the sensing information objective function calculation unit, a step of storing the current location information of all UAVs placed in the sensor field by the multiple UAV current position data storage unit and providing it to the sensing information objective function calculation unit, and PSO (Particle Swam Optimization) based overall objective function maximum UAV position calculation unit satisfies the constraints required for UAV network operation using the PSO algorithm and includes the step of deriving the position of the UAV that can maximize the calculated overall objective function. do.

According to embodiments of the present invention, information collected in a wide range of areas for various sensor operating environments, including environmental monitoring and surveillance, detection of events such as disaster situations, and movement tracking of specific objects using sensors in an outdoor environment, is transmitted to an unmanned aircraft and an unmanned air vehicle. It is obtained efficiently through communication with sensors and transmitted to a remote server using an unmanned aerial vehicle network. Through this, information from a wide range of sensor fields can be quickly acquired through unmanned aerial vehicles in various applications that require real-time response. In addition, in determining the optimal movement position of multiple UAVs asynchronously, the value of sensing information obtained from sensors is maximized, and at the same time, the time and energy required for acquiring and moving sensing information are comprehensively considered to increase efficiency in UAV operation. You can.

Figure 1 is a diagram showing the configuration of an optimal movement path and sensing information acquisition system using PSO-based biomimetic machine learning in sensor data acquisition using an unmanned flying vehicle according to an embodiment of the present invention.
Figure 2 is a flowchart illustrating a method for obtaining optimal movement path and sensing information using PSO-based biomimetic machine learning in sensor data acquisition using an unmanned flying vehicle according to an embodiment of the present invention.
Figure 3 is a diagram showing a grid-based 3D topographic map and model for the entire sensor area according to an embodiment of the present invention.
Figure 4 is a diagram showing UAV sensing coverage, which is a grid area of sensing data that an unmanned aircraft can obtain from sensors on the ground according to the position of the unmanned aircraft in three-dimensional space according to an embodiment of the present invention.
FIG. 5 is a diagram illustrating a process for performing steps to check whether a random grid cell on the ground is included in the sensing coverage of a specific UAV according to an embodiment of the present invention.
FIG. 6 is a diagram illustrating a process for performing steps to check whether a random grid cell on the ground is included in the sensing coverage of a specific UAV according to an embodiment of the present invention.
Figure 7 is a diagram showing a method of acquiring sensing information for each time slot and moving to the next optimal location for one UAV according to an embodiment of the present invention.
Figure 8 is a diagram showing the operation of a plurality of UAVs according to time slots in determining each optimal position according to an embodiment of the present invention.
Figure 9 is a diagram showing the network structure after determining the optimal location and moving the optimal location using each time unit PSO according to an embodiment of the present invention.

The present invention acquires sensing information from wireless sensors installed outdoors using an unmanned aerial vehicle (UAV), and its efficiency is achieved by comprehensively considering the sum of the sensor's sensing information value, information acquisition time, and energy required for acquisition. It relates to a method and system for deriving optimal 3D location information and movement paths that can maximize UAV operation time and UAV energy use.

According to an embodiment of the present invention, it must be taken into account that the area and communication quality of sensor nodes that can transmit sensing data through communication with the UAV vary depending on the location of each UAV in three-dimensional space, and the information has already been obtained by another UAV. The time elapsed since acquisition of previous information should also be considered. Additionally, in a situation where multiple UAVs move asynchronously, each UAV must be connected to the overall UAV network. UAV operation time for information acquisition must be minimized, and UAV energy consumption for information acquisition and movement to the next point must also be minimized. Deriving a real-time optimal UAV location and movement path that satisfies these multiple requirements requires a complex process and a long time.

The present invention seeks to provide a method and system for deriving the optimal position in real time while satisfying all required conditions by utilizing the biomimetic learning-based PSO (Particle Swam Optimization) technique. Hereinafter, embodiments of the present invention will be described in detail with reference to the attached drawings.

Figure 1 is a diagram showing the configuration of an optimal movement path and sensing information acquisition system using PSO-based biomimetic machine learning in sensor data acquisition using an unmanned flying vehicle according to an embodiment of the present invention.

The present invention relates to a method of calculating the movement path of the optimal location points of each UAV in acquiring information sensed by pre-installed sensors in an outdoor environment in real time using a plurality of unmanned aerial vehicles (UAVs) and transmitting it to a remote ground server. will be. More specifically, in a situation where the 3D terrain map prepared in advance and the arrangement of sensors are known, the server asynchronously derives the next optimal location for acquiring additional sensor data at the time the acquisition of sensing information from the current location of each UAV ends. It's about how to do it.

The objective function for determining the optimal location in the present invention is the sensing information value of sensors capable of communicating with the UAV at the designated next location, UAV movement time and information acquisition time to the next location, energy for UAV movement to the next location, and It is defined by considering the energy for obtaining information at the next location.

In an embodiment of the present invention, instead of performing a complete global inspection of all possible points in three-dimensional space, the PSO method, a biomimetic technology, is used to quickly calculate the next optimal position of the UAV that needs to move. In order to make the search faster and reduce the computational load, the present invention divides the (x,y) coordinates of the sensor field into grid (cell) areas of a predefined size, places sensors in grid units, and terrain height at the center of the grid area. Save information such as The combination of sensor field grids capable of communicating on the ground according to the location of the UAV is determined by the directional antenna angle of the UAV, the UAV location, the shape of the ground terrain, and the transmission power of the sensor and UAV.

In acquiring sensor data using an unmanned aerial vehicle, the optimal movement path and sensing information acquisition system using PSO-based biomimetic machine learning consists of a UAV coverage calculation unit 110, a sensing value objective function calculation unit 120, and a UAV operation time objective. Function calculation unit 130, UAV operation energy consumption objective function calculation unit 140, sensing value calculation unit 150 for each sensing grid, multiple UAV current position data storage unit 160, and PSO-based overall objective function maximum value UAV position Includes a calculation unit 170.

The UAV coverage calculation unit 110 according to an embodiment of the present invention uses information about the pre-stored sensor placement map, 3D terrain map, and 3D no-fly zone map to determine the UAV's next sensing information acquisition location candidates. In order to calculate the evaluation value of the overall objective function, the range of the sensor area grids that can communicate at the location next to the UAV is calculated.

The sensing value objective function calculation unit 120 according to an embodiment of the present invention calculates the value of sensing information that can be obtained from sensors within UAV coverage at a location next to the calculated UAV.

The sensing value objective function calculation unit 120 according to an embodiment of the present invention defines the sensing information value obtainable from all grid cells within the coverage of the UAV as a normalized sensing value objective function.

The UAV operating time objective function calculation unit 130 according to an embodiment of the present invention estimates the flight time required to move the UAV from the current location to the next location and the time required to acquire sensing information at the next location to calculate the objective function. Calculate the value.

The UAV operation time objective function calculation unit 130 according to an embodiment of the present invention calculates the operation time from the current location to the next location, the estimated movement time for moving to the next location, and the estimated movement time for moving to the next location and the operation time from the sensors in coverage while staying at the next location. The sum of the estimated times to acquire sensing information is defined as the normalized UAV operation time objective function.

The UAV operation energy consumption objective function calculation unit 140 according to an embodiment of the present invention estimates the energy required to move the UAV from the current location to the next location and the energy required to acquire sensing information at the next location to calculate the objective function. Calculate the value.

The UAV operation energy consumption objective function calculation unit 140 according to an embodiment of the present invention calculates the operation energy consumption for the next evaluation location from the current location, UAV movement flight energy for moving to the next location, and acquisition of sensing information at the next location. For this purpose, the UAV operational energy consumption objective function is defined as the normalized sum of UAV stationary flight energy for stationary flight and UAV communication energy for communication between the UAV and sensors to acquire sensing information at the next location.

The sensing value calculation unit 150 for each sensing grid according to an embodiment of the present invention calculates the change in sensing information value according to the sensor type for each sensing grid over time and provides the values required for calculation by the sensing information objective function calculation unit. .

The multiple UAV current location data storage unit 160 according to an embodiment of the present invention stores the current location information of all UAVs deployed in the sensor field and provides it to the sensing information objective function calculation unit.

The PSO-based overall objective function maximum value UAV position calculation unit 170 according to an embodiment of the present invention satisfies the constraints required for UAV network operation using the PSO (Particle Swam Optimization) algorithm, and the calculated overall objective function Derive the position of the UAV that can maximize .

The PSO-based overall objective function maximum value UAV position calculation unit 170 according to an embodiment of the present invention performs a search by applying the PSO algorithm to reduce the amount of calculation of the objective function and enable real-time implementation, and determines the movement of each UAV. When moving to the next optimal location after completing the acquisition of sensing data from the current location, the new moving location is selected as the optimal location by classifying the location that satisfies predetermined conditions.

Figure 2 is a flowchart illustrating a method for obtaining optimal movement path and sensing information using PSO-based biomimetic machine learning in sensor data acquisition using an unmanned flying vehicle according to an embodiment of the present invention.

In the proposed sensor data acquisition using an unmanned aerial vehicle, the optimal movement path and sensing information acquisition method using PSO-based biomimetic machine learning is a sensor deployment map with a UAV coverage calculation unit pre-stored, a 3D terrain map, and a 3D no-fly zone. Step (210) of calculating the range of communicable sensor area grids at the next location of the UAV to calculate the evaluation value of the overall objective function for the candidates for the next sensing information acquisition location of the UAV using information about the map (210), sensing value purpose A step (220) in which the function calculation unit calculates the value of sensing information obtainable from sensors within the UAV coverage at the next location of the calculated UAV, and the UAV operating time objective function calculation unit calculates the value required to move the UAV from the current location to the next location. Calculating an objective function value by estimating the flight time and the time required to acquire sensing information at the next location (230), the UAV operation energy consumption objective function calculation unit calculates the energy required to move the UAV from the current location to the next location. and calculating the objective function value by estimating the energy required to acquire sensing information at the next location (240), the sensing value calculation unit for each sensing grid calculates the change in sensing information value according to the sensor type for each sensing grid over time. Step 250 of calculating and providing a value necessary for the calculation of the sensing information objective function calculation unit, the multiple UAV current location data storage unit stores the current location information of all UAVs placed in the sensor field and provides it to the sensing information objective function calculation unit. Step 260 and the maximum overall objective function based on PSO (Particle Swam Optimization) UAV position calculation unit satisfies the constraints required for UAV network operation using the PSO algorithm and maximizes the calculated overall objective function. It includes a step 270 of deriving the location of .

In step 210, the UAV coverage calculation unit calculates the overall objective function for the next sensing information acquisition location candidates of the UAV using information about the sensor deployment map, 3D terrain map, and 3D no-fly zone map stored in advance. To calculate the evaluation value, the range of communicable sensor area grids at the location next to the UAV is calculated.

In step 220, the sensing value objective function calculation unit calculates the value of sensing information obtainable from sensors within UAV coverage at a location next to the calculated UAV.

The sensing value objective function calculation unit according to an embodiment of the present invention defines the sensing information value obtainable from all grid cells within the coverage of the UAV as a normalized sensing value objective function.

In step 230, the UAV operation time objective function calculation unit calculates the objective function value by estimating the flight time required to move the UAV from the current location to the next location and the time required to acquire sensing information at the next location.

The UAV operating time objective function calculation unit according to an embodiment of the present invention estimates the operating time from the current location to the next location, the estimated travel time for moving to the next location, and the acquisition of sensing information from sensors in coverage while staying at the next location. The sum of times is defined as the normalized UAV operation time objective function.

In step 240, the UAV operation energy consumption objective function calculation unit calculates the objective function value by estimating the energy required to move the UAV from the current location to the next location and the energy required to acquire sensing information at the next location.

The UAV operation energy consumption objective function calculation unit according to an embodiment of the present invention calculates the operation energy consumption for the next evaluation location from the current location, the UAV movement flight energy for moving to the next location, and the UAV movement flight energy for moving to the next location. The sum of UAV stationary flight energy for stationary flight and UAV communication energy for communication between the UAV and sensors to acquire sensing information at the next location is defined as a normalized UAV operation energy consumption objective function.

In step 250, the sensing value calculation unit for each sensing grid calculates the change in the sensing information value according to the sensor type for each sensing grid over time and provides the value necessary for the calculation of the sensing information objective function calculation unit.

In step 260, the multiple UAV current location data storage unit stores the current location information of all UAVs placed in the sensor field and provides it to the sensing information objective function calculation unit.

In step 270, the maximum overall objective function based on PSO (Particle Swam Optimization) UAV position calculation unit satisfies the constraints required for UAV network operation using the PSO algorithm and maximizes the calculated overall objective function. Derive the location of the UAV.

The PSO-based overall objective function maximum value UAV position calculation unit according to an embodiment of the present invention applies the PSO algorithm to reduce the amount of calculation of the objective function and enable real-time implementation, and performs a search in the current location according to the movement of each UAV. When moving to the next optimal location after completing sensing data acquisition, the new moving location is selected as the optimal location by classifying the location that satisfies predetermined conditions.

Figure 3 is a diagram showing a grid-based 3D topographic map and model for the entire sensor area according to an embodiment of the present invention.

Figure 3(a) shows a grid-based topographic map, and Figure 3(b) shows the three-dimensional coordinates of grid cells.

The remote server according to an embodiment of the present invention uses the topographic map of the sensor field to construct a grid-based topographic map as shown in FIG. 3(a). The entire sensor field 311 is is defined in the three-dimensional space of, and the sensor field of the sensor field The plane is It is divided into grid cells 312 with a unit size of . In the entire 3D sensor field space, the no-fly zone is determined in advance in 3D space. Center point of ith grid cell If z _i is the height of the terrain corresponding to (321), the three-dimensional coordinates of the ith grid cell of the grid-based terrain map are It is defined as (322). When each UAV stops at a specific location to acquire sensing information, it is positioned on the center point of one of the defined grids. This is to reduce the amount of calculations required to determine the optimal location, including the value of sensing information. Therefore, in an embodiment of the present invention, the next location for obtaining the next sensing information of the UAV is If it exists above, the location (323) of the kth UAV is It is defined as

Figure 4 is a diagram showing UAV sensing coverage, which is a grid area of sensing data that an unmanned aircraft can obtain from sensors on the ground according to the position of the unmanned aircraft in three-dimensional space according to an embodiment of the present invention.

An area where communication is possible with sensors on the ground is determined according to the next location of the UAV according to an embodiment of the present invention. In Figure 4, the kth UAV ( ) To obtain sensing information from sensors on the ground at a random location 410, a directional antenna is used to broadcast a sensing information transmission request message including the user's ID. At this time, the angle of the 3D directional antenna is It is defined as (420). In order to communicate with sensors on the ground, the distance between the UAV and the sensors on the ground is the maximum communication distance. It must be smaller than (430). is determined by equation (1).

Equation (1)

In an embodiment of the present invention, sensors use directional antennas. Additionally, it is assumed that communication between the UAV and sensors takes place in free space. is the maximum message transmission distance from the UAV to the sensor, is the maximum message transmission distance from the sensor to the UAV. is the transmit antenna gain, is the receiving antenna gain, is the wavelength of the signal, is the transmit power of the UAV, is the transmission power of the sensor, is the minimum reception power required to receive a message. Among the grid cells where sensors on the ground exist, A set of grid cells that can communicate with Coverage ( ) is defined as (440). UAV coverage is determined by the location of the UAV, UAV antenna angle, mutual communication distance according to transmission and reception power requirements, and whether LOS (Line of Sight) is supported between the UAV and sensor grid cells according to the terrain type of the ground. In the model of the present invention, it is assumed that if the UAV and LOS are not secured at a specific ground point due to the curvature of the ground terrain, communication quality is poor and communication cannot be performed normally.

FIG. 5 is a diagram illustrating a process for performing steps to check whether a random grid cell on the ground is included in the sensing coverage of a specific UAV according to an embodiment of the present invention.

Figure 5(a) is a diagram for explaining the process for calculating the maximum length of the bottom circle in the cone area, and Figure 5(b) is a diagram for explaining the process for calculating the UAV beam angle area.

The 3D position of When, in the present invention, a random grid cell on the ground go The process of checking whether something is in coverage is as follows.

In step 1 Check the maximum transmitting and receiving floor area. Grid cells for communication between UAVs and sensors and The distance between Equation (1) is It must be smaller than Directional antenna angle Maximum transmitting and receiving distance when using Maximum length of the bottom circle in the cone area within the bottom (511) is calculated using equation (2).

Equation (2)

In a plane, grid cells center point of class location of distance between (512) If it is bigger than It is located outside of coverage. Therefore, random grid cells go To be included in coverage, equation (3) must be satisfied. In the embodiment of Figure 5 does not satisfy equation (3) It is outside the coverage area.

Equation (3)

In step 2, check the maximum communication distance. Grid cells that passed stage 1 Check whether it is located within the maximum communication distance. and any cell that passed step 1 3D distance from Check whether (513) satisfies equation (4). grid cell if satisfied Is is within the coverage of.

Equation (4)

In step 3, the UAV beam angle area is inspected. Grid cells that passed step 2 Check whether it is located within the directional antenna angle. and any cell that passed step 2 height of the center of the ground in Maximum distance from UAV in antenna radiation area (521) is obtained using equation (5), and cell 3D distance from 522 street Check whether it is smaller than . Additionally, since the UAV's directional antenna points toward the ground, the grid cell The height should be lower than the UAV height. Therefore, equation (6) must be satisfied.

Equation (5)

Equation (6)

FIG. 6 is a diagram illustrating a process for performing steps to check whether a random grid cell on the ground is included in the sensing coverage of a specific UAV according to an embodiment of the present invention.

Check LOS in step 4. Grid passed step 3 go Check whether communication is possible without geographical obstacles on the shortest propagation path. As shown in Figure 6 in a plane (611) and Combination of all grid cells on line 613 connecting 612 with a straight line (set of direct path cells between and )(613) grid cell 614 This is inspected grid cells and Investigate whether it interferes with direct LOS communication with . grid cells and In the LOS straight line between grid cells in combination node point of commerce It is defined as is calculated using equation (7).

Equation (7)

grid cells in combination height of go If it is lower than the grid cell is a grid cell and It does not interfere with direct LOS communication. thus All grid cells in combination If equation (8) is satisfied for the grid cell Is It is judged to be within the coverage of .

Equation (8)

Figure 7 is a diagram showing a method of acquiring sensing information for each time slot and moving to the next optimal location for one UAV according to an embodiment of the present invention.

FIG. 7(a) is a diagram for explaining the acquisition of sensing information for each time slot for one UAV, and FIG. 7(b) is a diagram for explaining movement to an optimal location.

According to an embodiment of the present invention When is currently in the optimal position after moving m times (701), Stays at the current location and receives sensing information from all sensors within the sensing coverage area (702). The time it takes is It is defined as Before the remote server completes receiving information from sensors within its sensing coverage, the remote server By determining the next optimal sensing information acquisition location, Delivered to (703). moves to the (m+1)th optimal location received from the server (704). At this time, the time required to move the UAV is It is defined as At the (m+1)th optimal position (705) Again obtains sensing information from sensors within coverage.

Figure 8 is a diagram showing the operation of a plurality of UAVs according to time slots in determining each optimal position according to an embodiment of the present invention.

Figure 9 is a diagram showing the network structure after determining the optimal location and moving the optimal location using each time unit PSO according to an embodiment of the present invention.

As in the example of FIG. 8, the next optimal position for each UAV is determined asynchronously (810).

Referring to FIG. 9, UAVs must be able to connect to the existing network even after moving to the next location (812), and at least one UAV in the UAV network must be able to connect to a Ground Base Station (GBS) connected to a server on the ground.

In the present invention, a remote server defines an objective function to determine the next optimal location for each UAV and selects a location that maximizes the objective function. The objective function is defined as a sensing value objective function (fitness function), a UAV operation time objective function, and a UAV operation energy consumption objective function, and the overall objective function is defined by adding the weights of the three objective functions.

As an example of calculating the sensed value objective function, In determining the (m+1)th new optimal sensing information acquisition location at time t, the sensing value fitness function at a specific location is As shown in equation (9), It is defined as the normalized value of the sensing information value obtainable from all grid cells within the coverage of .

Equation (9)

Is It is a set of grid cells within the coverage of . In the calculation of Among the grid cells within the coverage, grid cells that are currently acquiring information by another UAV are excluded. is the time In grid cells sensor type is the value of the sensed value function. Is is the estimated travel time from the mth current location to the (m+1)th location where the objective function is evaluated. is calculated using equation (10). silver It is the distance between the mth current position and the (m+1)th evaluation position. is the movement speed of the UAV is the sensor type is a value function according to the number of sensors per grid. is a grid cell sensor type is the number of sensors. is the normalization value for normalizing the sensed value objective function. is defined as equation (11). is the size of the grid cell defined in advance.

Equation (10)

Equation (11)

In equation (9) The function is a sensed value function that changes over time. The sensing value function is a UAV grid cell sensor type This indicates that when sensing information is obtained from sensors in , the sensing value changes depending on when the same sensor type was previously obtained in the corresponding grid cell. Maximum sensing value by sensor type and minimum sensing value This is defined in the dictionary and the sensor type in that cell If sensing information of is transmitted, the value of information resulting from additional acquisition of the same type of sensing information from the same cell is the minimum sensing value decreases to the maximum sensing value over time. increases until. Sensing value function in an embodiment of the present invention Depending on the application, can be set in various ways, such as an exponential function, linear function, or logarithmic function.

In equation (9) Silver sensor type is a value function based on the number of sensors per grid cell. If the number of sensors for each sensor type within a grid cell is too small, the sensing value of each sensor increases, and conversely, if the number of sensors exceeds the threshold, the sensing value of each sensor decreases. The value function value according to the number of sensors for each sensor type is preset. class It is decided between Value function according to the number of sensors in an embodiment of the present invention can be set in various ways, such as an exponential function, linear function, or logarithmic function, depending on the application.

As an example of calculating the UAV operation time objective function, The operating time for the next (m+1)th location from the current mth location is the estimated travel time to move to the next location ( ) and the estimated time to acquire sensing information from sensors within coverage while staying at the next location. The sum is normalized and calculated using equation (12). Equation (13) is the value for normalization in calculating Equation (12).

Equation (12)

Equation (13)

At the (m+1)th position for evaluation Estimated time to receive sensing information from all sensors within coverage is all grid cells within the coverage as in equation (12) All sensor types number of sensors It is the sum of the time required to receive the sensing information packet. is the sensor type This is the average delay time required to receive one sensing information packet. This value is predetermined depending on the protocol implementation for communication between the UAV and the sensor. is a grid cell sensor type This is the average packet transmission error probability during packet transmission. silver grid cell at the (m+1)th position of It is estimated using the free space signal-to-noise ratio (SNR) estimated using the distance. In equations (12) and (13), The value is the average number of packet transmissions from a given packet error probability to a successful packet transmission. In equation (13) for normalization, is preset as the UAV operation time objective function normalization parameter. is the total number of UAVs operating in the sensor field. is defined as the average packet error probability estimated using the SNR at the average distance of the possible distance combinations for the UAV and sensor grid cells in the sensor field.

As an example of calculating the UAV operation energy consumption objective function, The operational energy consumption for the next (m+1)th evaluation position from the current mth position is the UAV movement flight energy in equation (14) for movement to the next position. , UAV stationary flight energy in Equation (15) for the UAV to fly stationary to acquire sensing information at the next location. , UAV communication energy in equation (16) for communication between UAV and sensors to acquire sensing information at the next location. The sum of is defined as the objective function of equation (17), which is normalized. Equation (18) is the value for normalization in calculating Equation (17).

Equation (14)

Equation (15)

Equation (16)

Equation (17)

Equation (18)

is the unit energy per second required for mobile flight of the UAV. is the unit energy per second required for stationary flight of the UAV. is the sensor type is the communication unit energy required per packet transmission. In equation (18) is preset as the UAV operation energy consumption objective function normalization parameter.

Ultimately, the overall objective function that determines the next movement location of the UAV is the sensing value objective function as shown in Equation (19). , UAV operation time objective function , UAV operation energy consumption objective function It is defined as the weighted sum of . The overall objective function aims to maximize the sensing value objective function and minimize the UAV operation time objective function and UAV operation energy consumption objective function. Weight value of each objective function in equation (19) is determined by the operator according to the sensor network application using UAV.

Equation (19)

For a UAV that needs to move from the current position to the next position, the position that maximizes the overall objective function of equation (19) is calculated in the three-dimensional space within the sensor field. In order to reduce the amount of calculation and enable real-time implementation, the PSO (Particle Swam Optimization) algorithm is applied to speed up the search. When moving to the next (m+1)th optimal location after completing the acquisition of sensing data from the current location according to the m-th movement of each UAV, the new movement location must necessarily satisfy the following conditions and must not satisfy the conditions below. Locations that do not work are excluded from consideration for optimal location selection:

Condition 1: The location of the UAV is a defined 3D sensor field. must be located within

Condition 2: UAV's Plane coordinates are predefined It must be located in the center of one of the planar grid cells.

Condition 3: The height of the UAV is It must be higher than the ground level of the grid cells in a flat position.

Condition 4: The 3D position of the UAV must not fall within a predefined no-fly zone.

Condition 5: If other UAVs are in a location where direct communication with the ground base station (GBS) to which the server is connected is impossible, the next location of the UAV that needs to move must be located within the range where communication with the ground base station is possible.

Condition 6: The location of the UAV must be within a range that allows direct communication with at least one UAV among other UAVs.

In an embodiment using the PSO of the present invention The current position according to the mth movement of and the number of PSO particles is When , particles for evaluating the entire objective function at the following positions of The UAV position and particle speed of the first iteration are respectively and It is defined as Particles in the PSO algorithm of The particle movement speed of the first iteration is calculated using equation (20).

Equation (20)

is the weight value, class is a random value between 0 and 1, class is the weight used for the particle best and global best. is an iteration particles until This is the particle maximum value that maximizes Equation (19) during the movement of . is an iteration It is the global maximum value that maximizes Equation (19) among all particle movements up to . particle The next position of is updated as equation (21).

Equation (21)

The next optimal movement position of the UAV using PSO is determined by the global maximum value after a predefined number of iterations.

The device described above may be implemented with hardware components, software components, and/or a combination of hardware components and software components. For example, devices and components described in embodiments may include, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate array (FPGA), etc. , may be implemented using one or more general-purpose or special-purpose computers, such as a programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions. A processing device may execute an operating system (OS) and one or more software applications that run on the operating system. Additionally, a processing device may access, store, manipulate, process, and generate data in response to the execution of software. For ease of understanding, a single processing device may be described as being used; however, those skilled in the art will understand that a processing device includes multiple processing elements and/or multiple types of processing elements. It can be seen that it may include. For example, a processing device may include a plurality of processors or one processor and one controller. Additionally, other processing configurations, such as parallel processors, are possible.

Software may include a computer program, code, instructions, or a combination of one or more of these, which may configure a processing unit to operate as desired, or may be processed independently or collectively. You can command the device. Software and/or data may be used on any type of machine, component, physical device, virtual equipment, computer storage medium or device to be interpreted by or to provide instructions or data to a processing device. It can be embodied in . Software may be distributed over networked computer systems and stored or executed in a distributed manner. Software and data may be stored on one or more computer-readable recording media.

The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc., singly or in combination. Program instructions recorded on the medium may be specially designed and configured for the embodiment or may be known and available to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical media such as CD-ROMs and DVDs, and magnetic media such as floptical disks. -Includes optical media (magneto-optical media) and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, flash memory, etc. Examples of program instructions include machine language code, such as that produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc.

As described above, although the embodiments have been described with limited examples and drawings, various modifications and variations can be made by those skilled in the art from the above description. For example, the described techniques are performed in a different order than the described method, and/or components of the described system, structure, device, circuit, etc. are combined or combined in a different form than the described method, or other components are used. Alternatively, appropriate results may be achieved even if substituted or substituted by an equivalent.

Therefore, other implementations, other embodiments, and equivalents of the claims also fall within the scope of the claims described below.

Claims (7)

Using information about the pre-stored sensor deployment map, 3D terrain map, and 3D no-fly zone map, communication is possible at the next location of the UAV to calculate the evaluation value of the overall objective function for the candidates for the UAV's next sensing information acquisition location. A UAV coverage calculation unit that calculates the range of sensor area grids;
A sensing value objective function calculation unit that calculates the value of sensing information obtainable from sensors within the UAV coverage at the location next to the calculated UAV - The value of the sensing information is determined by the type of sensor that obtained the sensing information and the type of sensor that obtained the previous sensing information. It refers to the efficiency of sensing information that changes depending on the sensing information acquisition environment, including the elapsed time since -;
A UAV operation time objective function calculation unit that calculates an evaluation value of the objective function by estimating the flight time required to move the UAV from the current location to the next location and the time required to acquire sensing information at the next location;
A UAV operation energy consumption objective function calculation unit that calculates the evaluation value of the objective function by estimating the energy required to move the UAV from the current location to the next location and the energy required to acquire sensing information at the next location;
Maximum PSO-based overall objective function that uses the PSO (Particle Swam Optimization) algorithm to derive the position of the UAV that satisfies predetermined constraints for UAV network operation and ensures that the evaluation value of the calculated overall objective function has the maximum value. Value UAV location calculation unit - Predetermined constraints for the UAV network operation include UAV operation time for acquiring sensing information, UAV operation energy for acquiring sensing information and moving to the next location -;
The elapsed time since acquiring the previous sensing information to be provided to the sensing value objective function calculation unit according to the range of the sensor area grids communicable at the location next to the UAV calculated through the UAV coverage calculation unit and for each sensing grid A sensing value calculation unit for each sensing grid that calculates the sensing value for each sensing grid according to the sensor type; and
An optimal movement path and sensing information acquisition system for an unmanned aerial vehicle, including a multiple UAV current location data storage unit that stores the current location information of all UAVs placed in the sensor field and provides the current location information to the sensing value objective function calculation unit. delete According to paragraph 1,
The sensing value objective function calculation unit defines a sensing value objective function to represent the value of sensing information obtainable from all grid cells within the coverage of the UAV as a normalized value,
The UAV operating time objective function calculation unit normalizes the operating time from the current location to the next location by calculating the sum of the estimated travel time for moving to the next location and the estimated time to acquire sensing information from sensors within coverage while staying at the next location. Define the UAV operation time objective function to express it as a value,
The UAV operation energy consumption objective function calculation unit calculates the operation energy consumption for the next evaluation location from the current location to the UAV movement flight energy for moving to the next location, and the UAV stop for the UAV to stop and fly to acquire sensing information at the next location. Flight energy, defined as a UAV operation energy consumption objective function to express the sum of UAV communication energy as a normalized value for communication between the UAV and sensors to obtain sensing information at the next location
Optimal movement path and sensing information acquisition system for unmanned aerial vehicles. According to paragraph 1,
The PSO-based overall objective function maximum value UAV position calculation unit,
In order to reduce the amount of calculation of the objective function and enable real-time implementation, the PSO algorithm is applied to conduct the search, and after completing the acquisition of sensing data at the current position according to the movement of each UAV, when moving to the next optimal position, the new moving position is sensed. The optimal location is selected so that the evaluation values of the value objective function, UAV operation time objective function, and UAV operation energy consumption objective function have the maximum value.
Optimal movement path and sensing information acquisition system for unmanned aerial vehicles. The UAV coverage calculation unit uses information about the pre-stored sensor deployment map, 3D terrain map, and 3D no-fly zone map to calculate the evaluation value of the entire objective function for the candidates for the UAV's next sensing information acquisition location. calculating a range of communicable sensor area grids at a location next to the UAV;
A step in which the sensing value objective function calculation unit calculates the value of sensing information obtainable from sensors within the UAV coverage at a location next to the calculated UAV - the value of the sensing information is determined by the type of sensor that obtained the sensing information and the type of sensor that obtained the previous sensing information. It refers to the efficiency of sensing information that changes depending on the sensing information acquisition environment, including the elapsed time since -;
A UAV operating time objective function calculation unit calculating an evaluation value of the objective function by estimating the flight time required to move the UAV from the current location to the next location and the time required to acquire sensing information at the next location;
A step where the UAV operation energy consumption objective function calculation unit calculates an evaluation value of the objective function by estimating the energy required to move the UAV from the current location to the next location and the energy required to acquire sensing information at the next location;
Elapsed time since the sensing value calculation unit for each sensing grid acquired the previous sensing information to provide to the sensing value objective function calculation unit according to the range of the communicationable sensor area grids at the location next to the UAV calculated through the UAV coverage calculation unit. Calculating the sensing value for each sensing grid according to time and sensor type for each sensing grid;
A plurality of UAV current location data storage units storing the current location information of all UAVs placed in the sensor field and providing the information to a sensing value objective function calculation unit; and
PSO (Particle Swam Optimization)-based overall objective function maximum value UAV position calculation unit satisfies predetermined constraints for UAV network operation using the PSO algorithm, and the evaluation value of the calculated overall objective function has the maximum value. Step of deriving the location - Predetermined constraints for the UAV network operation include UAV operation time for acquiring sensing information, UAV operation energy for acquiring sensing information and moving to the next location -
Method for obtaining optimal movement path and sensing information of unmanned aerial vehicle, including. According to clause 5,
The sensing value objective function is defined as an objective function to express the value of sensing information obtainable from all grid cells within the coverage of the UAV as a normalized value,
The UAV operating time objective function is the normalized sum of the operating time from the current location to the next location, the estimated travel time to move to the next location, and the estimated time to acquire sensing information from sensors in coverage while staying at the next location. Define an objective function to express it as a value,
The UAV operation energy consumption objective function is the operation energy consumption for the next evaluation location from the current location, the UAV moving flight energy for moving to the next location, and the UAV stationary flight energy for the UAV to fly stationary to acquire sensing information at the next location. , defined as an objective function to express the sum of UAV communication energy as a normalized value for communication between the UAV and sensors to obtain sensing information at the next location.
Optimal movement path and sensing information acquisition method for unmanned aerial vehicles. According to clause 5,
The PSO-based overall objective function maximum value UAV position calculation unit uses the PSO algorithm to derive the position of the UAV that satisfies predetermined constraints for UAV network operation and ensures that the evaluation value of the calculated overall objective function has the maximum value. The steps are:
In order to reduce the amount of calculation of the objective function and enable real-time implementation, the PSO algorithm is applied to conduct the search, and after completing the acquisition of sensing data at the current position according to the movement of each UAV, when moving to the next optimal position, the new moving position is sensed. The optimal location is selected so that the evaluation values of the value objective function, UAV operation time objective function, and UAV operation energy consumption objective function have the maximum value.
Optimal movement path and sensing information acquisition method for unmanned aerial vehicles.