Propulsion energy consumption optimization method for distributed electric propulsion unmanned aerial vehicle

Abstract

The invention discloses a distributed electric propulsion unmanned aerial vehicle propulsion energy consumption optimization method. The method comprises the following steps: respectively establishing a flight dynamics model, an aerodynamic model and an electric propeller model of a distributed electric propulsion unmanned aerial vehicle; based on each established model, selecting a rotating speed group parameter as a decision parameter, determining an optimization target, constructing an optimization model, and setting a value space of the decision parameter; according to an improved adaptive particle swarm optimization algorithm based on an opposite dynamic learning strategy, the constructed optimization model is optimized, and a rotating speed sequence formed by electric propulsion rotating speeds on the left side and the right side of the distributed electric propulsion unmanned aerial vehicle in each time period is solved; the rotating speed sequence is substituted into the optimization model, the corresponding total energy consumption is calculated, and whether the rotating speed sequence enables the system to accurately reach a specified target point or not is verified; if the result meets the constraint condition, determining that the rotating speed sequence is an effective solution; and the rotating speed sequence and the total energy consumption form a final energy consumption optimization scheme.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of electric unmanned aerial vehicles, and particularly relates to a distributed electric propulsion unmanned aerial vehicle propulsion energy consumption optimization method. BACKGROUND

[0002] The electric unmanned aerial vehicle is one of the main forms of global aviation technology development at present, and can realize potential benefits of energy saving of more than 60%, emission reduction of more than 90%, and noise reduction of more than 65%, and is one of the most advanced research fields of the aviation industry.

[0003] Cruise and yaw control is the basic flight control function of unmanned aerial vehicles to realize flight trajectory tracking. The conventional cruise and yaw control, as shown in Figure 1 , is mainly realized by using flight control surfaces, but the flight control surfaces are mechanical devices, and mechanical failures such as high-altitude icing, rudder jamming, and rudder fracture are prone to occur. In recent years, with the development of distributed electric propulsion technology, the distributed electric propulsion system has become a prominent trend of the development of electric unmanned aerial vehicles. Unlike unmanned aerial vehicles using centralized propulsion systems, unmanned aerial vehicles using distributed electric propulsion (DEP) systems will have thrust distributed along the fuselage. As shown in Figure 2 , such characteristics make the cruise and yaw control of the unmanned aerial vehicle have additional controllability, because the moment can be generated by individually adjusting the thrust output of each propeller. The propulsion motors are independent of each other, induced by traditional flight control surfaces, and are called electric cruise and yaw control. Unlike traditional control methods, the control object of power cruise and yaw control is the propeller in the DEP system. It has been proven that implementing power cruise and yaw control can reduce or even eliminate the area of flight control surfaces, thereby improving the maximum endurance capability of the unmanned aerial vehicle.

[0004] The DEP system is the key core technology of the electric unmanned aerial vehicle. Distributed electric propulsion is the main source of thrust, which is usually composed of several electric motors distributed on the wings or fuselage, and the energy of the driving motor comes from the on-board power supply, the capacity of the battery or the amount of fuel reserve directly affects the flight time and the range of the task execution. By designing a more energy-saving flight trajectory for the DEP unmanned aerial vehicle, it can perform a larger range of flight tasks under the constraint of on-board energy. However, no one has proposed a propulsion energy consumption model suitable for the power yaw characteristics of DEP unmanned aerial vehicles, and the existing propulsion energy consumption model does not fully consider the propulsion characteristics and the characteristics of the on-board power supply, and the obtained propulsion energy consumption minimum trajectory cannot be practically applied.

[0005] Current fixed-wing UAV propulsion energy consumption modeling methods can be divided into several categories according to the characteristics of the UAV: only relying on rudder yaw, combining rudder yaw and centralized control of the propulsion motor. In 2017, Rui Zhang's team from the National University of Singapore established a theoretical model of the two-dimensional propulsion energy consumption of a fixed-wing UAV as a function of flight speed and acceleration, and on this basis, they defined the communication energy efficiency of the UAV. This model is the first theoretical model that links the energy consumption of the UAV to the speed (including flight speed and direction) and acceleration. In 2019, the team established a propulsion energy consumption model for a rotor UAV flying horizontally and designed a two-dimensional trajectory for the propulsion energy and communication energy of the UAV. However, the above models cannot reflect the DEP UAV's electric yawing characteristics, so they cannot fully utilize the advantages of the DEP system.

[0006] UAV path planning techniques can be divided into two categories: the first representation technique is based on c-space and relies on cell decomposition, roadmaps, potential fields, and Voronoi diagrams; the second category is represented by algorithms based on coordinate and non-coordinate techniques, such as genetic algorithms, evolutionary algorithms, simulated annealing algorithms, and ant colony optimization. The path planning problem of a UAV is treated as an optimization problem to find the optimal solution among all possible solutions. Due to the complexity of the DEP UAV propulsion energy consumption model, which involves aerodynamics, flight dynamics, and related complex nonlinear constraints of electric propulsion, existing optimization algorithms have difficulty handling this optimization problem. SUMMARY

[0007] The purpose of the present application is to provide a distributed electric propulsion UAV propulsion energy consumption optimization method to overcome the optimization difficulty problem existing in the prior art.

[0008] To achieve the above-mentioned tasks, the present application adopts the following technical solutions: A distributed electric propulsion UAV propulsion energy consumption optimization method, comprising: An aerodynamic model, a flight dynamics model, and an electric propulsion model of the distributed electric propulsion UAV are established respectively, based on the established models, a rotation speed group parameter is selected as a decision parameter and an optimization target is determined, an optimization model is constructed, and a value space of the decision parameter is set; According to the improved adaptive particle swarm optimization algorithm based on the dynamic learning strategy of opposition, the constructed optimization model is optimized, and the rotation speed sequence composed of the left and right electric propulsion rotation speeds of the distributed electric propulsion UAV in each time period is solved. Substitute the rotational speed sequence into the optimization model, calculate the corresponding total energy consumption, and verify whether the rotational speed sequence can enable the system to accurately reach the specified target point. If the result meets the constraints, the rotational speed sequence is confirmed as a valid solution. The rotational speed sequence and the total energy consumption constitute the final energy consumption optimization scheme.

[0009] Furthermore, the rotational speed parameters include the initial / final position ( ), minimum / maximum flight speed ), maximum acceleration ( ), maximum yaw rate ( ), No. Maximum rotational speed of each thruster ( ) and the maximum speed change rate of the propeller ( ).

[0010] Furthermore, the construction of the optimization model is specifically expressed as follows: ; ; in, express Time of the first Each thruster rotation speed, This indicates the number of sampling points used for discrete sampling of flight duration; For the number of thrusters, and For different energy consumption coefficients, , , , , These represent the phase voltage, back electromotive force coefficient, and resistance of the thruster, respectively. This represents the discrete time step.

[0011] Furthermore, the constraints of the optimization model include equation constraints and inequality constraints:

[0012] in express Real-time location coordinates of the drone ; and They represent Real-time flight speed and yaw angle of the drone; and They represent Real-time flight acceleration and yaw rate of the drone; ; ; ; ; ; wherein represents the rate of change of the rotational speed of the th propeller at the moment t, is the minimum rotational speed of the th propeller.

[0013] Further, when solving the optimization model, the relationship between the rotational speeds of the left and right electric propellers (ωL(t) and ωR(t)) and the rotational speed of the th propeller ωp(t) is set as follows:

[0014] wherein, is the number of propellers.

[0015] Further, the improved adaptive particle swarm optimization algorithm based on the opposite dynamic learning strategy is specifically as follows: Step 3.1, set the population size, the maximum number of iterations ; the search dimension of the population is , the upper limit of the parameter is , the lower limit of the parameter is ; the reverse learning ratio is ; the initial learning factor is and ; Step 3.2, set each particle to be a group of parameter vectors composed of the rotational speeds of the left and right electric propellers (ωL(t) and ωR(t)) at all time steps; randomly initialize the position vector x of the particle and the velocity vector v of the particle , which are defined as: ; ; wherein represents the position vector of the th particle, corresponding to a complete set of and time series; is each dimension component of the position vector of the particle, each component corresponding to the value of or at a time step; is each dimension component of the velocity vector of the particle; Step 3.3, calculate the objective function of the total energy consumption of the electric propulsion system​​​ As fitness function, the fitness evaluation after initialization with fitness function, and according to the particle initial position, velocity and fitness value initialization particle history optimal solution And global optimal solution ; Step 3.4, adaptive inertia weight Balancing of global search and local development, and adaptive learning for cognitive balance of particles and global: ; ; Wherein, Indicates the inertia weight at the th iteration; , Indicates the learning factor at the th iteration; Step 3.5, fixed proportion of reverse learning, before each generation Optimal particle generates reverse solution: ; Wherein, Indicates the position of the th particle, The position of the corresponding reverse particle; Evaluate the reverse solution, calculate the fitness value of the Reverse solution, and The optimal particle is compared and sorted; in all the optimal particles and corresponding reverse particles participating in the sorting, the top Particle is selected as the better solution, and is combined into the population to participate in the next iteration; Step 3.6, update the velocity and position of the particle; when the boundary is exceeded, the reflection boundary treatment is adopted, and the update formula is as follows:

[0016]

[0017] Wherein, , The velocity and position of the th particle at the th iteration; , Random number uniformly distributed in [0,1]; Step 3.7, repeat steps 3.4-3.6 until the maximum iteration number After stopping optimization; Step 3.8, when the iteration reaches the maximum number of times, stop the optimization process, extract the global optimal solution The position vector of the corresponding particle; the position vector contains the specific values of the left and right electric propulsion rotation speed (ωL, ωR) of all time steps, that is, the rotation speed sequence of the left and right electric propulsion of the unmanned aerial vehicle in each time period.

[0018] A terminal device, comprising a processor, a memory and a computer program stored in the memory; when the processor executes the computer program, the distributed electric propulsion unmanned aerial vehicle propulsion energy consumption optimization method is realized.

[0019] A computer readable storage medium, the medium stores a computer program; when the computer program is executed by a processor, the distributed electric propulsion unmanned aerial vehicle propulsion energy consumption optimization method is realized.

[0020] Compared with the prior art, the present application has the following technical characteristics: The present application creatively uses the optimization model combined with the improved adaptive particle swarm optimization (PSO) algorithm, adopts the dynamic learning strategy based on the opposition, minimizes the propulsion energy consumption of the DEP unmanned aerial vehicle in the limited time range for performing the flight task, fully considers the electric yawing characteristics of the distributed electric propulsion unmanned aerial vehicle, constructs the distributed electric propulsion unmanned aerial vehicle propulsion energy consumption theoretical model to describe the function relationship between the propulsion energy consumption of the DEP unmanned aerial vehicle and the propeller rotation speed, and determines the propeller rotation speed through multiple calculations and algorithm optimization, so as to optimize the minimum energy consumption of the unmanned aerial vehicle corresponding to the flight trajectory; compared with the prior art, the present method has higher optimization efficiency, and can significantly reduce the energy consumption of the unmanned aerial vehicle. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1 is a rudder yawing control schematic diagram; Figure 2 is an electric yawing control schematic diagram; Figure 3a is Flight trajectory diagram of two propulsion energy consumption optimization methods under flight task; Figure 3b is Speed comparison diagram of two propulsion energy consumption optimization methods under flight task; Figure 3c is Left and right propulsion motor rotation speed diagram obtained by the optimization method under flight task; Figure 4a is Flight trajectory diagram of two propulsion energy consumption optimization methods under flight task; Figure 4b is ​The flight speed comparison chart of two propulsion energy consumption optimization methods under flight task; Figure 4c is The left and right side propulsion motor speed chart obtained by the optimization method proposed under flight task. DETAILED DESCRIPTION

[0022] The scheme is aimed at the novel propulsion structure of DEP fixed-wing unmanned aerial vehicle, and a propulsion energy consumption model conforming to power yaw is built. The characteristics of the propulsion motor and the onboard power supply are comprehensively considered to find a real and feasible flight trajectory with minimum propulsion energy consumption. The propulsion energy consumption optimization method of the distributed electric propulsion unmanned aerial vehicle provided by the application is used for designing and optimizing the propulsion energy consumption of the DEP unmanned aerial vehicle. The scheme firstly analyzes the flight dynamics model of the distributed electric propulsion unmanned aerial vehicle to obtain specific parameters affecting energy consumption and build various dynamics models; then the optimization model is constructed to determine the decision parameters and constraint conditions; then the algorithm is improved based on the particle swarm algorithm by introducing the dynamic learning strategy based on opposition; finally, the improved algorithm is used to solve the optimal related parameters to calculate the optimal path.

[0023] The propulsion energy consumption optimization method of the distributed electric propulsion unmanned aerial vehicle provided by the application plans a flight trajectory with minimum propulsion energy consumption that meets the flight task requirements, and the propulsion energy consumption of the distributed electric propulsion unmanned aerial vehicle can be minimized along the trajectory; the method comprises the following steps: Step 1, the flight dynamics model, the aerodynamics model and the electric propulsion model of the distributed electric propulsion unmanned aerial vehicle are respectively established; based on the established flight dynamics model, the aerodynamics model and the electric propulsion model, the rotation speed group parameters are selected as the decision parameters, the optimization target is determined, and the optimization model is constructed.

[0024] The flight dynamics model of the distributed electric propulsion unmanned aerial vehicle is: ; Wherein, is the mass of the unmanned aerial vehicle; is the flight speed of the unmanned aerial vehicle; the parameter superscript dot represents differentiation, and the number of dots represents the order of differentiation, such as represents the first-order differentiation of , and the same below; represents the total thrust generated by the propulsion, and represent the thrust generated by the left and right propulsion respectively; represents the aerodynamic drag; represents the moment of inertia of the yaw of the unmanned aerial vehicle; represents the yaw angle of the unmanned aerial vehicle; represents the yaw moment generated by the propulsion; This indicates the distance between the centers of gravity of the left and right thrusters of the drone.

[0025] The aerodynamic model of the distributed electric propulsion UAV is as follows: ; in, Indicates aerodynamic lift; It is air density; Wing area; and These represent the lift coefficient and drag coefficient, respectively. Zero lift-drag coefficient; For wingspan efficiency (typical values ​​are between 0.7 and 0.85), The aspect ratio is the ratio of the wingspan to its aerodynamic width. For the drone's gravity, It is the acceleration due to gravity. This is a correction factor.

[0026] The electric thruster model for a distributed electric propulsion UAV is as follows: ; in For the first The thrust of each propeller; It is air density; It is the propeller thrust coefficient; For the first The diameter of each thruster; For the first The rotational speed of each thruster; For the first The electrical power consumption of each thruster; It is the electromagnetic torque coefficient; It is a propulsion unit shaft current; , , These represent phase voltage, back electromotive force coefficient, and resistance, respectively. This represents the total energy consumption of the electric propulsion system. This is the upper limit of the integration time, i.e., the statistical duration of energy consumption; The number of thrusters; and For different energy consumption coefficients, where , .

[0027] Based on the distributed electric propulsion unmanned aerial vehicle propulsion energy consumption optimization design problem, the constraint conditions involve the complex coupling among flight dynamics, aerodynamics and electric propulsion characteristics, the parameters to be optimized are determined as the decision parameters, and the final energy consumption is selected as the optimization target, and the reasonable path reaching the target point and meeting the constraint conditions is finally calculated.

[0028] On the basis of the above-constructed flight dynamics model, aerodynamics model and electric propulsion model, the rotation speed group parameters are selected as the decision parameters in the scheme, and the total energy consumption of the electric propulsion system is the optimization target The minimum; The optimization model can be expressed as: ; ; In the formula represents the initial / final position of the unmanned aerial vehicle; represents the minimum / maximum flight speed of the unmanned aerial vehicle; represents the maximum acceleration of the unmanned aerial vehicle; represents the maximum yaw angle speed of the unmanned aerial vehicle; represents the maximum rotation speed of the th propulsion; represents the maximum rotation speed change rate of the th propulsion; These parameters are rotation speed group parameters.

[0029] The constraint conditions of the minimum optimization problem include equality constraints and inequality constraints: ; Among them, represents the two-dimensional horizontal position coordinates of the unmanned aerial vehicle at time ; represents the two-dimensional horizontal initial position coordinates of the unmanned aerial vehicle; and respectively represent the flight speed and yaw angle of the unmanned aerial vehicle at time ; and respectively represent the initial flight speed and initial yaw angle of the unmanned aerial vehicle.

[0030] ; ; ; ; ; To solve the optimization problem , the flight time is discretized, and the optimization problem and the constraint conditions are discretized as: ; in express Real-time location coordinates of the drone This represents the number of sampling points used to discretely sample the flight duration. Indicates the discrete time step; and They represent Real-time flight speed and yaw angle of the drone; and They represent Real-time flight acceleration and yaw rate of the drone.

[0031] ; ; ; ; ; In the formula and They represent The constant thruster speed and rate of change of speed, Indicates the first Minimum rotational speed of each thruster; Total energy consumption of electric propulsion system As the objective function, it can be discretized as: ; The discretized optimization model is as follows: ; ; Step 2: Set the value space of the decision parameters respectively.

[0032] Rotation speed parameters are selected as decision parameters, and real UAVs must meet many physical constraints when performing flight missions, including initial / final positions. ), minimum / maximum flight speed ), maximum acceleration ( ), maximum yaw rate ( ), Maximum speed of the thruster ( ) and the maximum speed change rate of the propeller ( ).

[0033] In one embodiment of the present invention: Set the initial state of the drone as , , and two different flight mission target points and ; the real UAV performing the flight mission needs to satisfy many physical constraints, including initial / final position , minimum / maximum flight speed , maximum acceleration , maximum yaw rate , maximum propeller rotation speed , and maximum propeller rotation speed change rate : ; ; ; ; .

[0034] Step 3, according to the improved adaptive particle swarm optimization algorithm based on the dynamic learning strategy of opposition, the constructed optimization model is optimized, and the rotation speed sequence composed of the left and right side electric propulsion rotation speeds of the distributed electric propulsion UAV in each time period is solved; wherein and are related as follows: ; In one possible implementation, the improved adaptive particle swarm optimization algorithm based on the dynamic learning strategy of opposition is as follows: Step 3.1, initialization process: set the population size to , the maximum number of iterations to ; the search dimension of the population is , the upper bound of the given optimization parameter is , the lower bound of the parameter is ; the reverse learning ratio is ; the initial learning factor is and .

[0035] Step 3.2, set each particle to be a group of parameter vectors composed of the left and right side electric propulsion rotation speeds of all time steps; randomly initialize the position vector and the velocity vector of the particle, defined as: ; ; wherein represents the position vector of the th particle, corresponding to a complete group of and time series; are the dimensional components of the particle position vector, each component corresponding to a time step or value; are the dimensional components of the particle velocity vector, each component corresponding to the rate of change of the position vector, used to control the update amplitude of the particle position.

[0036] Step 3.3, the objective function as the fitness function, using the fitness function for initialization after fitness evaluation, and according to the initial position, velocity and fitness value of the particle (i.e. the total energy consumption of the electric propulsion system) to initialize the particle historical optimal solution and the global optimal solution .

[0037] Step 3.4, adaptive inertia weight is used to balance global search and local development, and adaptive learning is used to balance the cognition of particles and global: ; ; wherein, indicates the inertia weight at the th iteration, with an initial value of 1; , indicates the learning factor at the th iteration; and are both set to 2; in early iterations, is larger, encouraging global search of particles; in later iterations linearly decreases, linearly increases, encouraging fine search and accelerating local development.

[0038] Step 3.5, fixed proportion reverse learning, the top optimal particles in each generation generate reverse solutions: ; wherein, indicates the position of the th particle, is the position of the corresponding reverse particle.

[0039] Evaluate the reverse solution, calculate the fitness value of the reverse solutions, and compare the fitness values of the top optimal particles and sort them; among all the optimal particles participating in the sorting and the corresponding reverse particles, select the topThe particle with the optimal solution is merged into the population to participate in the next iteration. Step 3.6, velocity and position of the particle are updated; when the boundary is crossed, the reflection boundary treatment is adopted, and the update formula is as follows:

[0040]

[0041] wherein, , is the velocity and position of the particle in the i th iteration; , is a random number uniformly distributed in the interval [0, 1].

[0042] Step 3.7, repeat steps 3.4-3.6 until the maximum number of iterations is reached after stopping the optimization.

[0043] Step 3.8, when the iteration reaches the maximum number of times, stop the optimization process, and extract the global optimal solution corresponding to the position vector of the particle; the position vector contains the specific values of the left and right electric propulsion rotation speeds ) at all time steps, i.e., the target left and right electric propulsion rotation speeds in each time period.

[0044] Step 4, substitute the rotation speed sequence obtained in step 3 into the optimization model to calculate the corresponding total energy consumption (i.e., the final output value of the objective function), and verify whether the rotation speed sequence can make the system accurately reach the specified target point; if the result satisfies the constraint condition, it is confirmed that the rotation speed sequence is a valid solution; then the rotation speed sequence and the calculated total energy consumption constitute the final energy consumption optimization scheme.

[0045] Embodiment: Referring to Figure 2 , the existing conventional fixed-wing unmanned aerial vehicle with rudder propulsion energy consumption model is selected as a comparison, and the initial state of the unmanned aerial vehicle is set to , , and two different flight task target points and , the optimization method proposed in the application is used for path planning. The real unmanned aerial vehicle needs to meet many physical constraints when performing a flight task, including initial / final position, minimum / maximum flight speed, maximum acceleration, maximum yaw rate, maximum propeller speed and maximum propeller speed change rate: ; ; ; ; . ​​

[0046] Table 1 Comparison table of propulsion energy consumption of two flight missions

[0047] Figures 3a to 4c It is shown that the method proposed in the application can effectively plan the flight trajectory of the DEP-UAV relying on electric cruise and yaw control to meet the constraint conditions, and the electric cruise and yaw control has an advantage over the traditional rudder yaw control. As can be seen from Table 1, compared with the traditional rudder yaw control method, the method significantly reduces the flight energy consumption under the premise of ensuring the completion of the task by optimizing the control strategy.

[0048] The above examples are only used to illustrate the technical solutions of the present application, but not limit it; although the present application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that the technical solutions recorded in the foregoing examples can still be modified, or some technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A method for optimizing the energy consumption of distributed electric propulsion drones, characterized in that, The application relates to a distributed electric propulsion unmanned aerial vehicle (UAV) propulsion energy consumption optimization method. A flight dynamics model, an aerodynamics model and an electric propulsion model of the distributed electric propulsion UAV are respectively established; based on the established models, a rotating speed group parameter is selected as a decision parameter, an optimization target is determined, an optimization model is constructed, and a value space of the decision parameter is set; An improved adaptive particle swarm optimization algorithm based on a dynamic learning strategy of opposition is used to optimize the constructed optimization model, and a rotating speed sequence composed of left and right electric propulsion rotating speeds of the distributed electric propulsion UAV in each time period is solved; The rotating speed sequence is substituted into the optimization model, corresponding total energy consumption is calculated, and whether the rotating speed sequence can make the system accurately reach a specified target point is verified; If the result meets a constraint condition, the rotating speed sequence is confirmed as an effective solution; the rotating speed sequence and the total energy consumption constitute a final energy consumption optimization scheme.

2. The method of claim 1, wherein, The speed group parameters include the initial / final position ( ), minimum / maximum flight speed ), maximum acceleration ( ), maximum yaw rate ( ), No. Maximum rotational speed of each thruster ( ) and the maximum speed change rate of the propeller ( ).

3. The method of claim 1, wherein, The optimization model is constructed, and the construction is specifically represented as follows: ； ； wherein, denotes the time instant the number of propeller revolutions, denotes the number of sampling points for discretely sampling the flight time; is the number of propellers, and are different energy consumption coefficients, , , , , denote the phase voltage, the back electromotive force coefficient and the resistance of the propeller, respectively; denotes the discrete time step.

4. The method of claim 1, wherein, Constraint conditions of the optimization model include formula constraints and inequality constraints: wherein denotes the position coordinates of the UAV at the time instant t, ; and denote the flight speed and the yaw angle of the UAV at the time instant t; and denote the flight acceleration and the yaw angular speed of the UAV at the time instant t; ； ； ； ； ； wherein represents the moment the rate of change of the speed of the the minimum speed of the th propeller.

5. The distributed electric propulsion UAV propulsion energy consumption optimization method of claim 1, wherein, When solving the optimization model, the electric propulsion speeds on the left and right sides of the distributed electric propulsion UAV are ( ) and the Each thruster rotation speed The relationship is set as follows: wherein, N is the number of propellers.

6. The distributed electric propulsion UAV propulsion energy consumption optimization method of claim 1, wherein, The improved adaptive particle swarm optimization algorithm based on the dynamic learning strategy of opposition is specifically as follows: Step 3.1, set population size, maximum number of iterations ; search dimension of population is ; upper bound of parameter is ; lower bound of parameter is ; inverse learning ratio ; initial learning factor and ; Step 3.2, set the rotational speed of each particle to be electrically propelled from the left and right sides of all time steps ( A set of parameter vectors; randomly initialize the particle position vectors. and velocity vector Defined as: ； ； in Indicates the first The position vectors of each particle correspond to a complete set of... and Time series; These are the components of the particle's position vector in each dimension, with each component corresponding to a specific time step. or Values; These are the components of the particle's velocity vector in each dimension; Step 3.

3. Calculate the objective function of total energy consumption of the electric propulsion system As the fitness function, the fitness evaluation after initialization of the fitness function is used, and the particle history optimal solution is initialized according to the particle initial position, speed and fitness value And the global optimal solution ; Step 3.4, using adaptive inertia weight Balancing global search and local exploitation, and adaptive learning for cognitive balance between particles and globality: ； ； in, Indicates the first Inertia weights in the next iteration; , Indicates the first The learning factor at the next iteration; Step 3.5, fixed-ratio inverse learning, select top-1 particle per generation Generate inverse solution from top-1 particle ； wherein, represents the position of the is the position of the corresponding anti-particle.​ evaluate the inverse solution, calculate the fitness value of this inverse solution, and compare and rank the fitness values of the said best particles and the said inverse solutions; select the top particles among all the best particles and the corresponding inverse solutions as the better solutions, and incorporate them into the population for the next iteration.​ In step 3.6, the speed and position of the particle are updated; when the particle is out of the boundary, a reflection boundary treatment is adopted, and the updating formula is as follows: wherein, , is the first iteration of the velocity, position of the th particle; , is a random number uniformly distributed in the interval [0, 1]. Step 3.

7. Repeat steps 3.4-3.6 until the maximum number of iterations is reached Post-optimization stop; Step 3.8: After the maximum number of iterations is reached, stop the optimization process and extract the global optimal solution. The corresponding particle position vector; this position vector includes the electric propulsion rotation speeds on both sides of all time steps ( The specific values ​​of ) constitute the rotational speed sequence, that is, the left and right electric propulsion speeds of the electric propulsion UAV in each time period.

7. A terminal device comprising a processor, a memory, and a computer program stored in the memory; characterized in that, When the computer program is executed by the processor, the distributed electric propulsion UAV propulsion energy consumption optimization method according to any one of claims 1-6 is realized.

8. A computer readable storage medium having stored therein a computer program; characterized in that, When the computer program is executed by the processor, the distributed electric propulsion UAV propulsion energy consumption optimization method according to any one of claims 1-6 is realized.

Images