An air-ground collaborative distribution route optimization method for unmanned aerial vehicles

Abstract

The application discloses an air-ground collaborative distribution route optimization method for unmanned aerial vehicles, and belongs to the field of unmanned aerial vehicle system automatic logistics distribution. Firstly, kinematics and dynamics modeling of unmanned aerial vehicles in an air-ground collaborative logistics distribution system is carried out; then, the distribution distance between each unmanned aerial vehicle and all unmanned vehicles, the flight energy consumption and the safety distance threshold between unmanned aerial vehicles are simultaneously taken as optimization targets to construct a multi-objective optimization function; subsequently, a particle swarm optimization algorithm is adopted to obtain optimal control signals at the current time; finally, the optimal control signals are input into the navigation control system of each unmanned aerial vehicle to generate optimal distribution routes in real time. The application provides a new idea for unmanned aerial vehicle cluster route planning in an air-ground collaborative distribution scenario, and introduces a flight energy consumption index into traditional unmanned aerial vehicle cluster route planning based on the shortest distribution distance, which can not only save energy consumption but also improve endurance.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of unmanned aerial vehicle automatic logistics distribution, and particularly relates to an air-ground collaborative distribution route optimization method for unmanned aerial vehicles. BACKGROUND

[0002] With the rapid development of e-commerce and intelligent logistics industry, unmanned aerial vehicle collaborative distribution technology has become a key breakthrough to improve logistics efficiency. Traditional unmanned aerial vehicle cluster route planning methods mainly focus on single target optimization, mainly based on the shortest distribution path principle for route design, rarely considering the influence of energy dynamic dissipation on the endurance of unmanned aerial vehicles, resulting in low energy utilization rate and insufficient task persistence in actual application. When dealing with air-ground collaborative distribution scenarios, existing technologies often plan paths for unmanned aerial vehicles and unmanned vehicles as independent units, lack of collaborative modeling of heterogeneous platform motion coupling characteristics, and are difficult to achieve global optimal decision in dynamic environment. Especially in complex urban scenarios, the safety distance control and multi-objective collaborative optimization problem of unmanned aerial vehicle cluster are more prominent. Traditional static planning algorithms are poor in real-time performance and weak in constraint processing ability, which easily leads to route conflict or obstacle avoidance failure. SUMMARY

[0003] In order to solve the problems existing in the prior art, the purpose of the present application is to provide an air-ground collaborative distribution route optimization method for unmanned aerial vehicles. First, the kinematics and dynamics of the unmanned aerial vehicles involved in the air-ground collaborative logistics distribution system are modeled. Then, the shortest distance from each unmanned aerial vehicle to all unmanned vehicles and the maximum difference between energy absorption and dissipation are taken as optimization objectives, and the safety distance threshold between unmanned aerial vehicles is taken as a penalty term, so as to construct a multi-objective optimization function. Next, the particle swarm optimization algorithm is used to obtain the optimal control signal at the current time. Finally, the optimal control signal is input into the unmanned aerial vehicle cluster navigation control system to generate the optimal distribution route in real time.

[0004] To achieve the above purpose, the present application adopts the following technical solutions:

[0005] An air-ground collaborative distribution route optimization method for unmanned aerial vehicles, the specific steps of which are as follows:

[0006] Step 1: Build kinematic model and dynamic model of each unmanned aerial vehicle;

[0007] Step 2: Based on the output information of the kinematic model and the real-time position information of the ground unmanned vehicle, build the distribution distance index item and the safety distance threshold index item;

[0008] Step three: based on the output information of the dynamic model, build flight energy consumption index item;

[0009] Step four: distribute weights to the distribution distance index item and the flight energy consumption index item, and add a safety distance threshold index item to form a multi-objective optimization function;

[0010] Step five: based on the particle swarm optimization algorithm, calculate the specific value of the multi-objective optimization function corresponding to different particles, and take the particle corresponding to the minimum value as the optimal yaw rate control signal;

[0011] Step six: input the optimal yaw rate control signal into the control system of the unmanned aerial vehicle, and the air-ground collaborative distribution route optimization task can be completed.

[0012] The kinematic model and dynamic model of each unmanned aerial vehicle are built as follows:

[0013] (1) In the North-East ground coordinate system, the first unmanned aerial vehicle flying in the specified height horizontal plane is described as follows:

[0014]

[0015] wherein, , respectively represent the horizontal and vertical coordinate values (both collectively referred to as plane position information) of the first unmanned aerial vehicle; and respectively represent the derivatives of and ; represents the flight speed of the first unmanned aerial vehicle; and respectively represent the heading angle and yaw rate of the first unmanned aerial vehicle. The main purpose of building the kinematic model of the unmanned aerial vehicle is to feed back the plane position and heading angle information of the first unmanned aerial vehicle at each sampling time in real time.

[0016] In order to ensure flight safety, improve stability and protect the actuator of the unmanned aerial vehicle, the change range of the flight speed and roll angle of the first unmanned aerial vehicle needs to be limited:

[0017]

[0018] wherein, the flight speed setting value and the roll angle setting value both represent positive real numbers greater than zero.

[0019] In addition, in order to facilitate the description of the incident angle of the sunlight irradiated on the unmanned aerial vehicle wing surface, the incident angle should also be calculated according to the sailing speed And the yaw angle velocity Solve the roll angle The specific solving formula is as follows:

[0020]

[0021] Wherein, G represents the acceleration of gravity, usually 9.8 ; Arc tan represents the arc tangent function.

[0022] (2) Compared with the energy consumption of the signal processing system and the attitude angle adjusting system, the energy consumed by the unmanned aerial vehicle in overcoming the aerodynamic resistance in the low-altitude environment occupies the main part. Therefore, the present application mainly constructs the dynamics model of the engine thrust And the aerodynamic resistance Of the unmanned aerial vehicle in the process of level flight:

[0023]

[0024] Wherein , , Atmospheric density, Oswald efficiency factor, aspect ratio of wing respectively; , , , And The wing area of the first Unmanned aerial vehicle, resistance coefficient, zero-lift resistance coefficient constant, lift coefficient and mass.

[0025] The output information based on the kinematics model and the real-time position information of the ground unmanned vehicle are used to construct the delivery distance index item and the safety distance threshold index item, and the specific method is as follows:

[0026] (1) The sum of the Euclidean distances between the first Unmanned aerial vehicle and all ground unmanned delivery vehicles (referred to as unmanned vehicles) Can be described by the following mathematical formula:

[0027]

[0028] Wherein, , Respectively represent the horizontal and longitudinal coordinate values of the first Unmanned aerial vehicle; , respectively represent the horizontal and vertical coordinate parameters (real-time position of the unmanned vehicle) of the first unmanned vehicle; represents the total number of unmanned vehicles operating in the delivery network. However, in actual delivery, the shorter the delivery distance, the better, that is, the smaller the , the better, which is just the opposite of the optimization direction of the optimization function.

[0029] (2) In order to optimize all optimization indicators in the direction of maximum value, the inverse of the Euclidean distance can be processed, that is, the delivery distance is defined as

[0030]

[0031] (3) Assuming that there are unmanned aerial vehicles in the air coordination delivery network. In order to ensure flight safety, the distance between these unmanned aerial vehicles should be greater than or equal to the specified minimum safety distance . Based on this minimum safety distance , the following safety distance threshold indicator item can be constructed:

[0032]

[0033] , wherein represents the penalty item given when the distance between the unmanned aerial vehicles is less than the safety distance. The penalty item is usually a large negative number, and the recommended value is in the range of -1000 to -2000.

[0034] The output information based on the dynamic model is used to construct a flight energy consumption indicator item, and the specific method is as follows:

[0035] (1) According to the dynamic model of the first unmanned aerial vehicle, the consumed power of the driving system of each unmanned aerial vehicle in the sampling period within a specified time range can be constructed:

[0036]

[0037] , wherein represents the efficiency factor of the propeller, represents the integral from zero time to Ts.

[0038] (2) The unmanned aerial vehicle involved in the present application can continuously absorb solar energy during flight because the outer surface of such a vehicle is covered with solar panels. As for how to construct the energy absorption power of the solar panels, the following steps can be taken:

[0039] ​1) Calculate the declination angle of the sun on the kth day:

[0040]

[0041] where, represents the kth day from January 1st.

[0042] 2) Calculate the hour angle of the sun on the kth day (unit: rad):

[0043]

[0044] where, represents the specific time on the kth day.

[0045] 3) Calculate the altitude angle of the sun at that time (unit: rad):

[0046]

[0047] where, represents the local latitude (unit: rad).

[0048] 4) Calculate the azimuth angle of the sun (unit: rad):

[0049]

[0050] 5) Calculate the incident angle of the sun's radiation on the solar panel mounted on the kth frame of the unmanned aerial vehicle (unit: rad):

[0051]

[0052] 6) Calculate the direct light beam and diffuse reflection on the earth's horizontal surface respectively:

[0053]

[0054] where, represents the depth of the direct light beam, represents the depth of the diffuse reflection light beam, represents a constant coefficient related to solar irradiance, represents a solar irradiance coefficient related to ground reflection, and G represents air mass ratio.

[0055] 7) Calculate the total solar irradiance received by the kth frame of the unmanned aerial vehicle:

[0056]

[0057] wherein 、 、 Ei, Ei, and Ei represent the direct light beam irradiance, the diffuse reflection irradiance, and the ground reflection irradiance, respectively, Ei represents the ground reflection coefficient.

[0058] 8) calculating the energy absorption power of the solar panel covered by each unmanned aerial vehicle body within a specified time range :

[0059]

[0060] wherein, Ei represents the conversion efficiency coefficient of the solar cell, Ei represents the integral from zero time to Ts.

[0061] In summary, the flight energy consumption index item is defined as the difference between the energy absorption power of the solar panel covered by the unmanned aerial vehicle body within a specified time and the flight energy consumption index item generated during its own flight process:

[0062]

[0063] The distribution distance index item and the flight energy consumption index item are assigned weights and added to the safety distance threshold index item to form a multi-objective optimization function, and the specific method is as follows:

[0064] (1) In order to make the optimal solution not deviate to the optimization target item with a larger order of magnitude, it is necessary to normalize the distribution distance index item and the flight energy consumption index item. The processing method of the present application is to normalize the following formula after calculating the optimization index item in each round of rolling time domain:

[0065]

[0066] wherein, and respectively represent the normalized distribution distance index item and the normalized flight energy consumption index item of the th unmanned aerial vehicle at the th sampling time; and respectively represent the distribution distance index item and the flight energy consumption index item of the th unmanned aerial vehicle at the th sampling time, Ei represents the total number of sampling times selected in each round of iterative optimization.

[0067] (2) define the Multi-objective optimization function at a sampling moment :

[0068]

[0069] wherein, denotes a weight parameter corresponding to the distribution distance index item, denotes a weight parameter corresponding to the flight energy consumption index item.

[0070] The particle swarm optimization algorithm is used to calculate specific values of the multi-objective optimization function corresponding to different particles, and the particle corresponding to the minimum value is taken as the optimal yaw rate control signal, and the specific method is as follows. It can be divided into the following 5 steps:

[0071] (1) According to the population number and the predicted time series length , a particle swarm position matrix of rows and columns and a particle swarm speed matrix satisfying normal distribution are randomly initialized:

[0072]

[0073] wherein denotes a random number uniformly distributed in the interval (0, 1), and denote the maximum value and the minimum value of the position of each particle, and denote the maximum value and the minimum value of the movement speed of each particle. Each particle in the particle swarm position matrix can be regarded as a yaw rate control signal to be optimized.

[0074] (2) The multi-objective optimization function value corresponding to each particle in the particle swarm position matrix is calculated, wherein .

[0075] (3) Enter the iterative optimization link: first, update the position and speed of the particle swarm; then, limit the amplitude of the particle swarm; finally, calculate the multi-objective optimization function value corresponding to each particle. .

[0076] (4) Determine whether the maximum number of iterations is reached; if yes, output the particle corresponding to the minimum value of the current multi-objective optimization function (the optimal yaw rate control signal), otherwise continue to execute step (3).

[0077] ​(5) In the above iteration optimization part, the weight dynamic attenuation strategy is used to determine the weight parameter of particle motion speed :

[0078]

[0079] Wherein The weight initial value is represented by w0, The weight attenuation amplitude is represented by w, The total iteration number is represented by N, The current iteration index number is represented by n.

[0080] Compared with the prior art, the present application has the following advantages:

[0081] 1) The present application proposes a particle swarm optimization air-ground collaborative distribution route optimization method in the field of unmanned aerial vehicle automated logistics distribution, which innovatively introduces the particle swarm optimization algorithm based on biological intelligence into the unmanned aerial vehicle route planning control framework, significantly improves the global search ability and the accuracy of the calculation cost function.

[0082] 2) The present application constructs a multi-objective optimization function, which comprehensively constructs different kinds of optimization objectives such as the shortest pickup distance from unmanned aerial vehicle to unmanned vehicle, the maximum energy absorption dissipation difference and the safety distance penalty term between unmanned aerial vehicles. This method not only optimizes the distribution efficiency, but also innovatively introduces the flight energy consumption index item, effectively balancing the contradiction between distribution efficiency and energy consumption management. BRIEF DESCRIPTION OF DRAWINGS

[0083] Figure 1 The present application is an air-ground collaborative distribution route optimization method for unmanned aerial vehicles.

[0084] Figure 2 The present application is a yaw angular velocity control signal generation module flow chart based on particle swarm optimization.

[0085] Figure 3 The present application is an air-ground collaborative distribution network route optimization graph under the existing distribution distance constraint.

[0086] Figure 4 The present application is an air-ground collaborative distribution network route optimization graph under the dual constraints of distribution distance and energy. DETAILED DESCRIPTION

[0087] As shown in the present application, an air-ground collaborative distribution route optimization method for unmanned aerial vehicles, comprising the following steps: Figure 1

[0088] (1) Build the kinematic model and dynamic model of each unmanned aerial vehicle;

[0089] ​(2) Based on the output information of the kinematic model and the real-time position information of the ground unmanned vehicle, a delivery distance index item and a safety distance threshold index item are constructed;

[0090] (3) Based on the output information of the dynamic model, a flight energy consumption index item is constructed;

[0091] (4) The delivery distance index item and the flight energy consumption index item are assigned weights and added to the safety distance threshold index item, thereby forming a multi-objective optimization function;

[0092] (5) Based on the particle swarm optimization algorithm, the specific values of the multi-objective optimization function corresponding to different particles are calculated, and the particle corresponding to the minimum value is taken as the optimal yaw rate control signal;

[0093] (6) The optimal yaw rate control signal is input into the control system of the unmanned aerial vehicle, thereby completing the air-ground collaborative delivery route optimization task.

[0094] Embodiment

[0095] 1. Build a kinematic model and a dynamic model for each unmanned aerial vehicle

[0096] (1) First, build the kinematic model of the unmanned aerial vehicle as follows:

[0097]

[0098] Where the speed of each unmanned aerial vehicle is , and the total number is 3. The initial position of each unmanned aerial vehicle is set to , , ; the initial heading angle of each unmanned aerial vehicle is set to , , .

[0099] In order to ensure flight safety, improve stability and protect the actuators of the unmanned aerial vehicle, the change range of the speed and roll angle of the first unmanned aerial vehicle needs to be limited:

[0100]

[0101] Where , .

[0102] (2) The invention mainly constructs a dynamic model of the unmanned aerial vehicle during level flight, with the engine thrust and the aerodynamic drag as the core:

[0103]

[0104] where atmospheric density Oswald efficiency factor aspect ratio of the wing wing area constant zero-lift drag coefficient mass .

[0105] 2. Build delivery distance and safety threshold index items:

[0106] (1) The sum of the Euclidean distance between each unmanned aerial vehicle and all ground unmanned delivery vehicles (referred to as unmanned vehicles) The Euclidean distance can be described by the following mathematical formula:

[0107] where the total number of unmanned vehicles operating in the delivery network

[0108] The position information of these unmanned vehicles is described by the following time-dependent equation:

[0109]

[0110] (2) In order to optimize all optimization indicators in the direction of maximum value, the Euclidean distance can be processed by taking the reciprocal, i.e. the delivery distance is defined as

[0111]

[0112] (3) Assuming that there are unmanned aerial vehicles in the air-ground collaborative delivery network. In order to ensure flight safety, the distance between these unmanned aerial vehicles should be greater than or equal to the specified minimum safety distance Based on this minimum safety distance , the following safety distance threshold index item can be constructed:

[0113]

[0114] where represents the penalty term for unmanned aerial vehicles less than the safety distance, and the penalty term for exceeding the safety distance .

[0115] 3. Build flight energy consumption index items

[0116] (1) Build the power consumption of each unmanned aerial vehicle

[0117] ​

[0118] where the efficiency factor of the propeller , the sampling period .

[0119] (2) The energy absorption power of the solar panel is constructed, which can be divided into the following steps:

[0120] Step 1: Study the dynamic evolution of the air-ground cooperative unmanned flow distribution network at 12 o'clock on the 202nd day from January 1. Therefore, the declination angle parameter of the day is calculated as

[0121]

[0122] where .

[0123] Step 2: Calculate the hour angle (unit: rad):

[0124]

[0125] where .

[0126] Step 3: Calculate the altitude angle of the sun at this time (unit: rad):

[0127]

[0128] where .

[0129] Step 4: Calculate the azimuth angle of the sun (unit: rad):

[0130]

[0131] Step 5: Calculate the incident angle (unit: rad) of the sun's radiation to the solar panel carried on the wing of the unmanned aerial vehicle:

[0132]

[0133] Step 6: Calculate the direct light beam and the diffuse reflection on the earth's horizontal plane respectively:

[0134]

[0135] where the direct light beam depth , the diffuse reflection light beam depth , and the constant coefficient of solar irradiance .

[0136] ​Step 7: Calculate the total solar irradiance received by each unmanned aerial vehicle : :

[0137]

[0138] where the ground reflection coefficient .

[0139] Step 8: Calculate the energy absorption power of the solar panels covered by each unmanned aerial vehicle body within a specified time range :

[0140]

[0141] where the conversion efficiency coefficient of the solar cell .

[0142] In summary, the flight energy consumption index item is defined as the difference between the energy absorption power of the solar panels covered by the unmanned aerial vehicle body within a specified time range and the flight energy consumption index item generated during its own flight process:

[0143]

[0144] 4. Constructing a multi-objective optimization function

[0145] (1) First, design the following normalization processing formula:

[0146]

[0147] where and represent the normalized delivery distance index item and the normalized flight energy consumption index item of the th unmanned aerial vehicle at the th sampling time, respectively; and represent the delivery distance index item and the flight energy consumption index item of the th unmanned aerial vehicle at the th sampling time, respectively, denotes the total number of sampling times selected for each iteration optimization.

[0148] (2) Define the multi-objective optimization function at the th sampling time :

[0149]

[0150] where , .

[0151] 5. Obtain the optimal yaw rate control signal.

[0152] As shown in Figure 2 , the following five steps can be divided:

[0153] Step 1: According to the population number =40 and the predicted time series length , randomly initialize the particle swarm position matrix and the particle swarm velocity matrix satisfying the normal distribution:

[0154]

[0155] Wherein, the maximum value and the minimum value of the position of each particle, the maximum value and the minimum value of the movement speed of each particle.

[0156] Step 2: Calculate the multi-objective optimization function value corresponding to each particle in the particle swarm position matrix , wherein .

[0157] Step 3: Enter the iterative optimization link: first, update the position and speed of the particle swarm. Then, limit the amplitude of these particle swarms. Finally, calculate the multi-objective optimization function value corresponding to each particle .

[0158] Step 4: Judge whether the maximum iteration number is reached; if yes, output the particle corresponding to the minimum value of the current multi-objective optimization function (the optimal yaw rate control signal), otherwise continue to execute step (3).

[0159] Step 5: In the above iterative optimization link, the weight parameter of the particle movement speed is determined by using the following weight dynamic attenuation strategy:

[0160]

[0161] Wherein, the initial weight value , the weight attenuation amplitude , the total iteration number , represents the current iteration index number.

[0162] ​​​​​In order to prove the superiority of the application in energy saving, comparative simulation is made in the embodiment. Figure 3 The traditional unmanned aerial vehicle cluster distribution route operation trajectory obtained by optimizing from the aspects of distribution distance and safety is shown, wherein 3 unmanned aerial vehicles store energy 1.4663 KJ in addition to the energy consumed for providing power for themselves. However, the simulation result as shown in the following table is obtained by introducing the flight energy consumption index term and designing reasonable weight between the flight energy consumption index term and the distance index term. Figure 4 According to statistical analysis, it can be found that after comprehensively considering the flight energy consumption and the distribution distance, the unmanned aerial vehicle cluster stores energy 1.5252 KJ in 50 seconds of distribution time, and the energy storage efficiency is improved by 4%.

[0163] The application deeply fuses the air-ground cooperative route planning and the intelligent particle swarm optimization algorithm, and establishes a multi-dimensional cooperative optimization system for the cooperative distribution distance, flight energy consumption and safety distance threshold of unmanned vehicle-unmanned aerial vehicle. The application combines the efficient optimization characteristics of the particle swarm algorithm, and effectively solves the deficiencies of the traditional method in multi-objective cooperation, dynamic environment adaptation and real-time control and the like. The technical scheme of the application not only improves the energy utilization efficiency of the air-ground cooperative logistics distribution system, but also significantly enhances the task reliability in the complex environment through dynamic route optimization, and provides an innovative solution for the efficient operation of the air-ground cooperative logistics distribution system.

Claims

1. A route optimization method for air-ground collaborative delivery of unmanned aerial vehicles, characterized by: The steps include: Step 1: Build the kinematic and dynamic models of each UAV; Step 2: Based on the output information of the kinematic model and the real-time location information of the ground unmanned vehicle, construct the delivery distance indicator and the safety distance threshold indicator; Step 3: Construct flight energy consumption indicators based on the output information of the dynamic model; Step 4: Assign weights to the delivery distance index and the flight energy consumption index and add the safety distance threshold index to form a multi-objective optimization function; Step 5: Based on the particle swarm optimization algorithm, calculate the specific values ​​of the multi-objective optimization function corresponding to different particles, and use the particle corresponding to the minimum value as the optimal yaw rate control signal; Step 6: Input the optimal yaw rate control signal into the navigation control system of each unmanned aerial vehicle, thus completing the air-ground collaborative delivery route optimization task; The output information of the kinematic model and the real-time position information of the ground unmanned aerial vehicle are used to construct the delivery distance index item and the safety threshold index item. The specific method is as follows: first, the sum of the distances between the i-th unmanned aerial vehicle and the real-time positions of all unmanned aerial vehicles is calculated; then, the inverse of the sum of the distances is used as the delivery distance index item; then, it is determined whether the distance between the unmanned aerial vehicles exceeds a given threshold. If so, a negative number with an absolute value in the range of 1000-2000 is designed as the safety distance threshold index item.

2. The method for optimizing air-ground collaborative delivery routes for unmanned aerial vehicles according to claim 1, characterized in that: Build the kinematic model and dynamic model of each unmanned aerial vehicle. The specific method is as follows: first, the speed V of the i-th unmanned aerial vehicle is a,i and yaw angular velocity ω a,i As control input, a kinematic model is constructed that can provide real-time feedback of plane position information, heading angle, and roll angle; among them, the roll angle φ a,i Using the formula φ a,i =arctan((V a,i ω a,i ) / g) is solved indirectly, where g is the gravitational acceleration constant; then a dynamic model considering engine thrust and aerodynamic drag is constructed.

3. The method for optimizing air-ground collaborative delivery routes for unmanned aerial vehicles according to claim 2, characterized in that: The kinematic model is as follows: Among them, x a,i 、y a,i Respectively represent the horizontal and vertical coordinate values ​​of the i-th unmanned aerial vehicle; and Represents x a,i and y a,i The derivative of V a,i represents the speed of the i-th unmanned aerial vehicle; ψ i and ω a,i denote the heading angle and yaw rate of the i-th unmanned aerial vehicle respectively; Considering the engine thrust T i and aerodynamic drag F x,i The kinetic model is as follows: T i =F x,i Among them ρ, ε, R a They represent the atmospheric density, Oswald efficiency factor, and the aspect ratio of the wing respectively; S i 、C x,i 、 C z,i and m i They represent the wing area, drag coefficient, zero-lift drag coefficient constant, lift coefficient and mass of the i-th unmanned aerial vehicle respectively.

4. The method for optimizing air-ground collaborative delivery routes for unmanned aerial vehicles according to claim 2, wherein: Speed ​​V a,i The absolute value of the speed is less than or equal to the speed setting value Roll angle φ a,i The absolute value of the roll angle is less than or equal to the set value 5. The method for optimizing air-ground collaborative delivery routes for unmanned aerial vehicles according to claim 1, characterized in that: The output information of the dynamic model is used to construct a flight energy consumption index. The specific method is as follows: first, the power consumption of the drive system of each unmanned aerial vehicle is calculated; then, the energy absorption power of the solar panels covering the fuselage of each unmanned aerial vehicle is calculated; and finally, the difference between the energy absorption power and the energy consumption power of each unmanned aerial vehicle is used as the flight energy consumption index.

6. The method for optimizing air-ground collaborative delivery routes for unmanned aerial vehicles according to claim 1, characterized in that: Weights are assigned to the delivery distance index and the flight energy consumption index, and a safety distance threshold index is added to form a multi-objective optimization function. The specific method is as follows: first, the delivery distance index and the flight energy consumption index corresponding to the d-th sampling moment and the total number of moments in the future are calculated and normalized, where D represents the length of the predicted time series; then, weight parameters with a sum of 1 are assigned to the normalized delivery distance index and the flight energy consumption index, and a safety distance threshold index is added to form the final multi-objective optimization function.

7. The method for optimizing air-ground collaborative delivery routes for unmanned aerial vehicles according to claim 6, characterized in that: The method for calculating the delivery distance index item and the flight energy consumption index item corresponding to the d-th sampling moment and its future total D-1 moments and performing normalization processing is: Normalized delivery distance index item where R Φ,i,d represents the delivery distance index item of the i-th unmanned aerial vehicle at the d-th sampling time; the normalized flight energy consumption index item where R P,i,d Represents the flight energy consumption index item of the i-th unmanned aerial vehicle at the d-th sampling moment.

8. The method for optimizing air-ground collaborative delivery routes for unmanned aerial vehicles according to claim 1, characterized in that: Based on the particle swarm optimization algorithm, the specific values ​​of the multi-objective optimization function corresponding to different particles are calculated, and the particle corresponding to the minimum value is used as the optimal yaw rate control signal. The specific method is as follows: first, a particle swarm is randomly generated according to the population size and the length of the predicted time series. Next, a weighted dynamic attenuation strategy is used to iteratively optimize the particle swarm. Finally, the particle corresponding to the minimum value of the multi-objective optimization function at the current moment is used as the optimal yaw rate control signal for a certain unmanned aerial vehicle.

9. The method for optimizing air-ground collaborative delivery routes for unmanned aerial vehicles according to claim 8, characterized in that: Weight dynamic attenuation strategy, its specific calculation formula is W e =W I -(W S e) / E all , where W e The weight parameter representing the particle speed, W I Represents the initial value of weight, W S Represents the weight attenuation amplitude, E all Represents the total number of iterations, and e represents the current iteration index.

Images