A method and system for unmanned aerial vehicle trajectory planning and release decision based on model predictive control

Abstract

The application provides a model prediction control-based unmanned carrier aircraft trajectory planning and launching decision method and system, and belongs to the field of unmanned aircraft trajectory planning. In order to solve the problems that the existing sub-aircraft launching timing decision problem is difficult to be mathematically modeled and is difficult to be combined with the carrier aircraft flight trajectory planning, the application combines the carrier aircraft trajectory planning task and the sub-aircraft launching task, constructs a cost function related to the carrier aircraft flight and the sub-aircraft launching state according to the task requirement, models the carrier aircraft flight and the sub-aircraft launching in a prediction time domain as a mixed integer programming problem, and solves the optimal flight trajectory of the carrier aircraft and the launching instruction of the sub-aircraft through rolling optimization to generate the optimal flight trajectory of the carrier aircraft and the launching instruction of the sub-aircraft. The application establishes a solving framework for the combined unmanned aircraft sub-aircraft launching decision problem, unifies the trajectory planning problem and the launching decision problem, and improves the autonomy and intelligence of the unmanned aircraft.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) trajectory planning technology, and more specifically, to a method and system for UAV trajectory planning and deployment decision-making based on model predictive control. Background Technology

[0002] Combined unmanned aerial vehicles (UAVs) are a new type of aircraft that consists of a carrier UAV carrying several unmanned sub-UAVs. Traditional small UAVs are small in size, highly agile, and have high maneuverability, but they have poor endurance and low environmental perception accuracy, making it impossible to obtain precise environmental information. On the other hand, large UAVs have a large payload and long endurance, and can be equipped with multiple sensors to obtain precise local information, but they have poor maneuverability and stealth, and cannot operate in remote and narrow environments. Combined UAVs combine the advantages of both, and their configuration design can well meet the actual mission requirements.

[0003] Unmanned aerial vehicle (UAV) trajectory planning and deployment decision-making refers to the process during flight where the UAV, based on environmental changes, mission requirements, and safety considerations, uses intelligent algorithms to select appropriate flight maneuvers and paths, and deploys or releases sub-units at suitable times based on mission requirements. Current UAV trajectory planning and deployment decision-making methods can be broadly categorized into two types: AI-based methods and optimization theory-based methods. AI-based methods mainly include rule-based expert system methods and deep neural network-based reinforcement learning methods. Rule-based expert system methods rely on an expert system library built based on IF-THEN rules, performing corresponding maneuvers or actions when the rules are met. However, building an expert system requires a large sample database. Unlike expert systems, deep neural network-based reinforcement learning methods do not require samples and can achieve autonomous decision-making through self-learning and self-updating. However, existing reinforcement learning methods face challenges such as training complexity, stability issues, and security concerns, making it difficult to guarantee effectiveness. Optimization theory-based methods mostly model the problem as a global optimization problem of flight trajectory, often yielding a globally optimal solution. However, the adaptability of global optimization algorithms needs improvement in highly dynamic and uncertain application scenarios. For example, the existing patent "Integrated Method and System for Deployment Decision and Path Planning of Mother-Daughter UAVs" (application number: CN202211403755.7) constructs a 3D map environment based on the mission requirements of the mother-daughter UAVs; determines the number of daughter UAVs and the target positions of each daughter UAV; establishes a path cost function based on the target positions and terrain altitudes of each daughter UAV; and uses an improved particle swarm optimization algorithm to obtain the flight paths of the mother platform and daughter UAVs in the mother-daughter UAVs, where the initial flight path point of the daughter UAV is the deployment position. Focusing on the selection of the separation point unique to the mother-daughter UAV deployment process, this method fills a research gap in path planning methods for mother-daughter UAVs. Addressing the problem that traditional particle swarm optimization algorithms converge too quickly and are prone to getting trapped in local optima, an improved particle swarm optimization algorithm is proposed, which balances local and global optima in path optimization, effectively improving the quality of the planned path.

[0004] Model predictive control (MDI) is an optimal control method where the current control action is obtained by solving an open-loop optimal control problem in a finite time domain online at each sampling time. The system only implements the first action of the obtained optimal control sequence onto the controlled object. This rolling optimization strategy allows MDI to flexibly adjust according to the actual situation at different times, thus exhibiting strong adaptability and flexibility, and better adapting to highly dynamic and uncertain real-world application environments. Simultaneously, MDI can naturally handle multivariate control problems, coordinating the relationships between variables through optimized control strategies to achieve overall optimal control. However, current applications of MDI in UAVs are mostly concentrated in trajectory planning and tracking, lacking research on the decision-making problem of combined UAV deployments. Summary of the Invention

[0005] The technical problem to be solved by this invention is:

[0006] To address the difficulty in mathematical modeling for the current decision-making process regarding the timing of aircraft deployment, and the challenge in integrating it with the flight trajectory planning of the carrier aircraft.

[0007] The technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows:

[0008] This invention provides a method for unmanned aerial vehicle trajectory planning and deployment decision-making based on model predictive control, comprising the following steps:

[0009] S100. Establish the mission environment of the carrier aircraft, including establishing a discrete kinematic model of the unmanned carrier aircraft, determining the number of carrier aircraft, determining the control input of the carrier aircraft, setting the release decision variables, determining the initial position information and threat capability of the target, and determining the flight start and end point of the carrier aircraft.

[0010] S200, Predicting the target trajectory, including establishing a motion model of the dynamic target, given the prediction time domain of the model prediction control of the dynamic target, and predicting the motion trajectory of the target in a prediction time domain based on the motion model of the dynamic target.

[0011] S300. Establish a mixed integer programming problem for single prediction time domain downloader trajectory planning and sub-aircraft deployment decision, including the prediction time domain of model predictive control of a given carrier aircraft, and establish a cost function for carrier trajectory planning and sub-aircraft deployment decision. The cost function includes a weighted sum of path cost, deployment cost, threat cost and altitude cost.

[0012] S400, Solve the mixed integer programming problem established in step S300, and use the first value of the solved carrier control input and the submachine deployment decision variable as the control input and deployment decision variable at the next sampling time.

[0013] S500. Repeat steps S200, S300 and S400. After all the sub-units are deployed, set the weight of the path cost in the cost function to 1 and the weight of the deployment cost to 0. When the carrier aircraft flies to the vicinity of the destination, the mission is completed.

[0014] Furthermore, step S100 specifically includes:

[0015] The mission environment of the carrier aircraft is established, and the kinematic model of the carrier aircraft is as follows:

[0016]

[0017] Where x, y, z represent the coordinates of the aircraft in the inertial frame; ψ represents the yaw angle of the aircraft; γ is the track angle of the aircraft; v is the velocity of the aircraft, which is set to a constant; r is the yaw rate of the aircraft; and q is the pitch rate of the aircraft.

[0018] Taking the sampling time as T, after discretization, we have:

[0019]

[0020] Where i represents the i-th time after discretization;

[0021] Assume that the delivery only occurs at the discrete sampling points, and let the delivery decision variable be u. drop The value can be either 0 or 1, where 0 means no delivery and 1 means delivery.

[0022] The threat posed by the target to the carrier aircraft is quantified using the following model:

[0023]

[0024] Among them, P threat R represents the probability that the aircraft is threatened. threat d represents the threat radius; d is the straight-line distance between the aircraft and the threat target; outside the threat radius, the probability of the aircraft being threatened is 0.

[0025] Furthermore, step S200 specifically includes:

[0026] Establish the following dynamic objective model:

[0027]

[0028] Where, x t ,y t v represents the target's position coordinates; tx ,v ty Let a be the velocity of the target in the x and y directions; tx ,a ty Let x be the acceleration of the target in the x and y directions.

[0029] Based on the observations obtained by the sensors on the aircraft, the position coordinates of the target within a time domain are predicted step by step.

[0030] Furthermore, step S300 specifically includes:

[0031] Construct a cost function for carrier trajectory planning and sub-aircraft deployment decision-making, assuming the prediction time domain is N. p The cost function is:

[0032] J = w1C path +w2C drop +w3Crisk +w4C height (5)

[0033] Among them, C path C drop C risk C height These represent path cost, deployment cost, threat cost, and altitude cost, respectively; w1, w2, w3, and w4 are the corresponding weights; the specific expressions for each cost are as follows:

[0034] S310. Path cost: Within a prediction time domain, the sum of the straight-line distances from each trajectory point to the destination is defined as the path cost within that prediction time domain, expressed as follows:

[0035]

[0036] Where, N p For the prediction time domain in model predictive control problems; D end D is the straight-line distance from the aircraft to the destination. max w1 represents the maximum range of the carrier aircraft; when the carrier aircraft has not yet completed its deployment mission, w1 = 0, and after deployment, w1 > 0.

[0037] S320, Deployment Cost, using a deployment decision variable u for each sampling point in the prediction time domain. drop (i) The product of the delivery cost corresponding to the sampling point is used as the delivery cost of the sampling point, if u drop If (i) is 0, then the deployment cost at that point is 0. The specific expression is as follows:

[0038]

[0039] Among them, C R Cost of delivery distance; C A Cost of the angle of deployment; w A This represents the weight of the deployment angle cost relative to the deployment distance cost.

[0040] S330, Threat Cost: Within a prediction time domain, the sum of probabilities of the aircraft being threatened is set as the threat cost within that prediction time domain, expressed as:

[0041]

[0042] Among them, P threatj This represents the threat cost corresponding to the j-th target, and n represents the total number of targets;

[0043] S340, Height Cost, the specific expression is as follows:

[0044]

[0045] in,

[0046]

[0047] Among them, h min This is the minimum altitude of the aircraft.

[0048] Furthermore, in step S320, the specific expressions for the delivery distance cost and angle cost are as follows:

[0049] S321. Distance cost of deployment: Set the target's maximum range d within the sub-aircraft. max Within this range, a minimum delivery distance d is simultaneously set. min The expression for the distance cost of delivery is as follows:

[0050]

[0051] Where d is the straight-line distance between the carrier aircraft and the target;

[0052] S322, Deployment Angle Cost: Define the line of sight between the aircraft and the target as the line of sight, and the angle between the aircraft speed and the line of sight as the forward angle;

[0053] The cost of setting the launch angle of the carrier aircraft is as follows:

[0054] C A =(1-A H / π)(1-A V / π) (11)

[0055] Among them, A H A is the angle between the aircraft's horizontal speed and the projection of the line of sight onto the flight plane. V The angle between the longitudinal velocity of the aircraft and the projection of the line of sight onto the longitudinal plane.

[0056] Furthermore, step S300 also includes:

[0057] S350, Non-integer constraints, setting upper and lower bound constraints for the aircraft's yaw angle ψ, track angle γ, yaw rate r, and pitch rate q:

[0058] -π≤ψ≤π

[0059] γ min ≤γ≤γ max

[0060] -r max <r<r max

[0061] -q max ≤q≤q max

[0062] S360, Integer term constraint, requires each submachine to perform one delivery within a prediction time domain, expressed by the formula:

[0063]

[0064] In summary, the problem of carrier trajectory planning and sub-aircraft deployment decision-making can be expressed as:

[0065]

[0066] In the index function, we set w1 to 0 or 1, w2 to 2, w3 to 1, and w4 to 1.

[0067] S400, Solve the mixed integer programming problem established in step S300, and use the first value of the solved carrier control input and the submachine deployment decision variable as the control input and deployment decision variable at the next sampling time.

[0068] S500. Repeat steps S200, S300 and S400. After all the sub-units are deployed, set the weight of the path cost in the cost function to 1 and the weight of the deployment cost to 0. When the carrier aircraft flies to the vicinity of the destination, the mission is completed.

[0069] A model predictive control-based unmanned aerial vehicle (UAV) trajectory planning and deployment decision-making system is provided. The system has program modules corresponding to the above steps and executes the steps in the model predictive control-based UAV trajectory planning and deployment decision-making method during runtime.

[0070] A computer-readable storage medium storing a computer program configured to, when invoked by a processor, implement steps of a model predictive control-based unmanned aerial vehicle trajectory planning and deployment decision-making method.

[0071] Compared with the prior art, the beneficial effects of the present invention are:

[0072] This invention addresses the difficulty in mathematical modeling the sub-aircraft deployment timing decision problem and its inability to integrate it with the carrier aircraft flight trajectory planning problem. It proposes an integrated rolling optimization framework based on model predictive control, which unifies the sub-aircraft deployment decision problem and the carrier aircraft trajectory planning problem into a mixed integer programming problem. The carrier aircraft control input and sub-aircraft deployment decision are obtained at each sampling time through rolling optimization.

[0073] This invention proposes a cost function for trajectory planning and deployment decision-making. It comprehensively considers the impact of the distance and angle relationship between the carrier and the target on the deployment decision, while also taking into account the safety of the carrier. Ultimately, it guides the carrier to fly to the preset endpoint, improving the intelligence and autonomy of the combined UAV. Compared with existing methods, it has better adaptability and stability, a wide range of applications, and high engineering application value. Attached Figure Description

[0074] Figure 1 This is a flowchart of a model predictive control-based unmanned aerial vehicle trajectory planning and deployment decision-making method in an embodiment of the present invention;

[0075] Figure 2 This is a graph showing the changes in the input variables yaw angle and track angle in an embodiment of the present invention;

[0076] Figure 3 This is a diagram showing the flight trajectory of the carrier aircraft and the trajectory of the daughter aircraft deployment in an embodiment of the present invention. Detailed Implementation

[0077] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0078] Specific Implementation Plan 1: Combining Figure 1 As shown, this invention provides a method for unmanned aerial vehicle trajectory planning and deployment decision-making based on model predictive control, including the following steps:

[0079] S100. Establish the mission environment of the carrier aircraft, including establishing a discrete kinematic model of the unmanned carrier aircraft, determining the number of carrier aircraft, determining the control input of the carrier aircraft, setting the release decision variables, determining the initial position information and threat capability of the target, and determining the flight start and end point of the carrier aircraft.

[0080] Specifically, it includes:

[0081] The mission environment of the carrier aircraft is established, and the kinematic model of the carrier aircraft is as follows:

[0082]

[0083] Where x, y, z represent the coordinates of the aircraft in the inertial frame; ψ represents the yaw angle of the aircraft; γ is the track angle of the aircraft; v is the velocity of the aircraft, which is set to a constant value here; r is the yaw rate of the aircraft; and q is the pitch rate of the aircraft.

[0084] Taking the sampling time as T, after discretization, we have:

[0085]

[0086] Here, i represents the i-th time point after discretization; to facilitate subsequent bombing decisions, it is assumed that bombing only occurs at the discretized sampling points, and the bombing decision variable is denoted as u. drop The value is limited to 0 or 1, where 0 means no deployment and 1 means deployment. In the subsequent optimization problem, the yaw rate r, pitch rate q, and deployment decision variable u of the aircraft are considered. drop This will be used as a variable to be optimized;

[0087] The threat posed by the target to the carrier aircraft is quantified using the following model:

[0088]

[0089] Among them, P threat R represents the probability that the aircraft is "threatened"; threat Represents the threat radius; d is the straight-line distance between the aircraft and the threat target;

[0090] Outside the threat radius, the probability of the aircraft being threatened is 0; inside the threat radius, the probability of the aircraft being threatened gradually increases as d decreases.

[0091] S200, Predicting the target trajectory, including establishing a motion model of the dynamic target, given the prediction time domain of the model prediction control of the dynamic target, and predicting the motion trajectory of the target in a prediction time domain based on the motion model of the dynamic target.

[0092] Specifically, it includes:

[0093] Establish the following dynamic objective model:

[0094]

[0095] Where, x t ,y t v represents the target's position coordinates; tx ,v ty Let a be the velocity of the target in the x and y directions; tx ,a ty Let x be the acceleration of the target in the x and y directions.

[0096] Based on the observations measured by the sensors on the aircraft, the position coordinates of the target within a time domain are predicted step by step;

[0097] S300. Establish a mixed integer programming problem for single-prediction-time domain carrier trajectory planning and sub-carrier deployment decision-making, including the prediction time domain of the given carrier's model predictive control. Establish a cost function for carrier trajectory planning and sub-carrier deployment decision-making. The cost function is a weighted sum of path cost, deployment cost, threat cost, and altitude cost. The path cost is the sum of the distances from the carrier's trajectory points to the endpoint within a prediction time domain. The deployment cost is related to the carrier's distance and angle relative to the target. The threat cost is related to the carrier's distance relative to the target. It is stipulated that the weight of the path cost is 0 before all sub-carrier deployments are completed. Set motion constraints for the carrier in the mission, stipulating that the sum of the deployment decision variables within a prediction time domain must be 0.

[0098] Specifically, it includes:

[0099] First, we construct the cost function for carrier trajectory planning and sub-aircraft deployment decision-making, assuming the prediction time domain is N. p The design includes a cost function consisting of four parts:

[0100] J = w1C path +w2C drop +w3C risk +w4C height (5)

[0101] Among them, C path C drop C risk C height These represent path cost, deployment cost, threat cost, and altitude cost, respectively; w1, w2, w3, and w4 are the corresponding weights; the specific expressions for each cost are as follows:

[0102] S310. Path cost: Within a prediction time domain, the sum of the straight-line distances from each trajectory point to the destination is defined as the path cost within that prediction time domain, expressed as follows:

[0103]

[0104] Where, N p For the prediction time domain in model predictive control problems; D end D is the straight-line distance from the aircraft to the destination. max w1 represents the maximum range of the carrier aircraft; when the carrier aircraft has not yet completed its deployment mission, w1 = 0, and after deployment, w1 > 0.

[0105] S320. Deployment Cost: To accurately represent the impact of different deployment timings on the indicator function, this invention uses a deployment decision variable u for each sampling point within the prediction time domain. drop (i) The product of the delivery cost corresponding to the sampling point is used as the delivery cost of the sampling point, if u dropIf (i) is 0, then the deployment cost at that point is 0. The specific expression is as follows:

[0106]

[0107] Among them, C R Cost of delivery distance; C A Cost of the angle of deployment; w A This represents the weight of the deployment angle cost relative to the deployment distance cost; the specific expressions for deployment distance cost and angle cost are as follows:

[0108] S321. Distance cost of deployment: When deploying, ensure the target is within the maximum range d of the sub-aircraft. max Within this range, and considering that the sub-unit needs to undergo operations such as folding and unfolding its wings after deployment, a minimum deployment distance d is set here. min The expression for the distance cost of delivery is as follows:

[0109]

[0110] Where d is the straight-line distance between the carrier aircraft and the target;

[0111] S322, Cost of Launch Angle: After launch, the aircraft needs to adjust its heading and track angles to engage the target as quickly as possible; the line of sight between the aircraft and the target is defined as the line of sight, and the angle between the aircraft's speed and the line of sight is defined as the lead angle; to facilitate rapid adjustment after launch, the lead angle of the aircraft before launch needs to be as small as possible.

[0112] Therefore, the cost of setting the launch angle of the carrier aircraft is as follows:

[0113] C A =(1-A H / π)(1-A V / π) (9)

[0114] Among them, A H A is the angle between the aircraft's horizontal speed and the projection of the line of sight onto the flight plane. V The angle between the longitudinal velocity of the aircraft and the projection of the line of sight onto the longitudinal plane;

[0115] S330, Threat Cost: Within a prediction time domain, the sum of probabilities that the aircraft is "threatened" is set as the threat cost within that prediction time domain, expressed as:

[0116]

[0117] Among them, P threatj This represents the threat cost corresponding to the j-th target, and n represents the total number of targets;

[0118] S340, Altitude Cost: The probability of the carrier aircraft crashing or being shot down often increases as the flight altitude decreases. Furthermore, low-altitude deployments are detrimental to the attitude adjustment of the sub-aircraft. Therefore, this paper defines an altitude cost term C. height To avoid the aircraft flying too low, the specific expression is as follows:

[0119]

[0120] in,

[0121]

[0122] Among them, h min This is the minimum altitude of the aircraft.

[0123] S350. Non-integer constraints: For ease of subsequent problem solving, only upper and lower bound constraints are set for the yaw angle ψ, track angle γ, yaw rate r, and pitch rate q of the aircraft.

[0124] -π≤ψ≤π

[0125] γ min ≤γ≤γ max

[0126] -r < r < r

[0127] max max

[0128] -q max ≤q≤q max

[0129] S360. Integer term constraint: Model predictive control only uses the first element of the solved control input as the actual control input. Therefore, within a prediction time domain, solutions other than the first sampling time have no impact on the actual deployment decision. For ease of solution, it is mandatory that each slave machine must be deployed once within a prediction time domain, expressed by the formula:

[0130]

[0131] In summary, the problem of carrier trajectory planning and sub-aircraft deployment decision-making can be expressed as:

[0132]

[0133] The weights for each item in the indicator function are set as follows:

[0134] Table 1. Weight Setting of Index Function

[0135]

[0136] S400, Solve the mixed integer programming problem established in step S300, and use the first value of the solved carrier control input and the submachine deployment decision variable as the control input and deployment decision variable at the next sampling time.

[0137] S500. Repeat steps S200, S300 and S400. After all the sub-units are deployed, set the weight of the path cost in the cost function to 1 and the weight of the deployment cost to 0. When the carrier aircraft flies to the vicinity of the destination, the mission is completed.

[0138] The other combinations and connections in this implementation scheme are the same as in Specific Implementation Scheme 1.

[0139] Simulation Experiment

[0140] With the flight start point of the carrier aircraft set at (0,0,30) and the flight end point at (100,50,30), use (x) end ,y end ,z end This indicates that the initial yaw angle of the aircraft is 0.3 rad, and the initial track angle is 0 rad. Assume there is a dynamic ground target in the environment with initial coordinates (20, 70, 0), initial velocity (4.777, 1.478, 0), and initial acceleration (0.070, -0.243, 0). The aircraft can acquire the target's position, velocity, and acceleration in real time through sensors. In step S300, let N... p It is 5, D max Let d be 1000; max =10, d min =50; h min The value is 20; in step S350, let:

[0141] γ min =γ max =30°

[0142] r max =45° / s

[0143] q max =10° / s

[0144] The flight path and drop point of the carrier aircraft, such as Figure 3 As shown, the black solid line represents the trajectory of the carrier aircraft, the blue solid line represents the trajectory of the moving target, the blue pentagram represents the final position of the moving target, the blue solid dot represents the drop point of the sub-aircraft, the red dashed line represents the trajectory of the sub-aircraft, the blue dashed circle represents the target threat range, and the blue square represents the destination of the carrier aircraft's flight. As can be seen, the carrier aircraft completed its drop mission, ultimately flying to the destination without being threatened by the target. Figure 2As shown, the carrier aircraft's trajectory is smooth and meets the constraints. This verifies the effectiveness of the proposed carrier aircraft trajectory planning and sub-aircraft deployment decision algorithm, demonstrating promising prospects for engineering applications.

[0145] While the present invention has been disclosed above, its scope of protection is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention, and all such changes and modifications will fall within the scope of protection of the present invention.

Claims

1. A method for trajectory planning and deployment decision-making of unmanned aerial vehicles based on model predictive control, characterized in that, Includes the following steps: S100. Establish the mission environment of the carrier aircraft, including establishing a discrete kinematic model of the unmanned carrier aircraft, determining the number of carrier aircraft, determining the control input of the carrier aircraft, setting the release decision variables, determining the initial position information and threat capability of the target, and determining the flight start and end point of the carrier aircraft. S200, Predicting the target trajectory, including establishing a motion model of the dynamic target, given the prediction time domain of the model prediction control of the dynamic target, and predicting the motion trajectory of the target in a prediction time domain based on the motion model of the dynamic target. S300. Establish a mixed integer programming problem for single prediction time domain downloader trajectory planning and sub-aircraft deployment decision, including the prediction time domain of model predictive control of a given carrier aircraft, and establish a cost function for carrier trajectory planning and sub-aircraft deployment decision. The cost function includes a weighted sum of path cost, deployment cost, threat cost and altitude cost. S400, Solve the mixed integer programming problem established in step S300, and use the first value of the solved carrier control input and the submachine deployment decision variable as the control input and deployment decision variable at the next sampling time. S500. Repeat steps S200, S300 and S400. After all the sub-units are deployed, set the weight of the path cost in the cost function to 1 and the weight of the deployment cost to 0. When the carrier aircraft flies to the vicinity of the destination, the mission is completed.

2. The method for unmanned aerial vehicle trajectory planning and deployment decision-making based on model predictive control according to claim 1, characterized in that, Step S100 specifically includes: The mission environment of the carrier aircraft is established, and the kinematic model of the carrier aircraft is as follows: Where x, y, z represent the coordinates of the aircraft in the inertial frame; ψ represents the yaw angle of the aircraft; γ is the track angle of the aircraft; v is the velocity of the aircraft, which is set to a constant; r is the yaw rate of the aircraft; and q is the pitch rate of the aircraft. Taking the sampling time as T, after discretization, we have: Where i represents the i-th time after discretization; Assume that the delivery only occurs at the discrete sampling points, and let the delivery decision variable be u. drop The value can be either 0 or 1, where 0 means no delivery and 1 means delivery. The threat posed by the target to the carrier aircraft is quantified using the following model: Among them, P threat R represents the probability that the aircraft is threatened. threat d represents the threat radius; d is the straight-line distance between the aircraft and the threat target; outside the threat radius, the probability of the aircraft being threatened is 0.

3. The method for unmanned aerial vehicle trajectory planning and deployment decision-making based on model predictive control according to claim 2, characterized in that, Step S200 specifically includes: Establish the following dynamic objective model: Where, x t ,y t v represents the target's position coordinates; tx ,v ty Let a be the velocity of the target in the x and y directions; tx ,a ty Let x be the acceleration of the target in the x and y directions. Based on the observations obtained by the sensors on the aircraft, the position coordinates of the target within a time domain are predicted step by step.

4. The method for unmanned aerial vehicle trajectory planning and deployment decision-making based on model predictive control according to claim 3, characterized in that, Step S300 specifically includes: Construct a cost function for carrier trajectory planning and sub-aircraft deployment decision-making, assuming the prediction time domain is N. p The cost function is: J=w1C path +w2C drop +w3C risk +w4C height (5) Among them, C path C drop C risk C height These represent path cost, deployment cost, threat cost, and altitude cost, respectively; w1, w2, w3, and w4 are the corresponding weights; the specific expressions for each cost are as follows: S310. Path cost: Within a prediction time domain, the sum of the straight-line distances from each trajectory point to the destination is defined as the path cost within that prediction time domain, expressed as follows: Where, N p For the prediction time domain in model predictive control problems; D end D is the straight-line distance from the aircraft to the destination. max w1 represents the maximum range of the carrier aircraft; when the carrier aircraft has not yet completed its deployment mission, w1 = 0, and after deployment, w1 > 0. S320, Deployment Cost, using a deployment decision variable u for each sampling point in the prediction time domain. drop (i) The product of the delivery cost corresponding to the sampling point is used as the delivery cost of the sampling point, if u drop If (i) is 0, then the deployment cost at that point is 0. The specific expression is as follows: Among them, C R Cost of delivery distance; C A Cost of the angle of deployment; w A This represents the weight of the deployment angle cost relative to the deployment distance cost. S330, Threat Cost: Within a prediction time domain, the sum of probabilities of the aircraft being threatened is set as the threat cost within that prediction time domain, expressed as: in, This represents the threat cost corresponding to the j-th target, and n represents the total number of targets; S340, Height Cost, the specific expression is as follows: in, Among them, h min This is the minimum altitude of the aircraft.

5. The method for unmanned aerial vehicle trajectory planning and deployment decision-making based on model predictive control according to claim 4, characterized in that, In step S320, the specific expressions for the delivery distance cost and angle cost are as follows: S321. Distance cost of deployment: Set the target's maximum range d within the sub-aircraft. max Within this range, a minimum delivery distance d is simultaneously set. min The expression for the distance cost of delivery is as follows: Where d is the straight-line distance between the carrier aircraft and the target; S322, Deployment Angle Cost: Define the line of sight between the aircraft and the target as the line of sight, and the angle between the aircraft speed and the line of sight as the forward angle; The cost of setting the launch angle of the carrier aircraft is as follows: C A =(1-A H / π)(1-A V / p) (11) Among them, A H A is the angle between the aircraft's horizontal speed and the projection of the line of sight onto the flight plane. V The angle between the longitudinal velocity of the aircraft and the projection of the line of sight onto the longitudinal plane.

6. The method for unmanned aerial vehicle trajectory planning and deployment decision-making based on model predictive control according to claim 5, characterized in that, Step S300 further includes: S350, Non-integer constraints, setting upper and lower bound constraints for the aircraft's yaw angle ψ, track angle γ, yaw rate r, and pitch rate q: -π≤ψ≤π c min ≤γ≤γ max -r＜r＜r max max -q max ≤q≤q max S360, Integer term constraint, requires each submachine to perform one delivery within a prediction time domain, expressed by the formula: In summary, the problem of carrier trajectory planning and sub-aircraft deployment decision-making can be expressed as:

7. The method for unmanned aerial vehicle trajectory planning and deployment decision-making based on model predictive control according to claim 6, characterized in that: In the index function, we set w1 to 0 or 1, w2 to 2, w3 to 1, and w4 to 1.

8. A model predictive control-based unmanned aerial vehicle trajectory planning and deployment decision-making system, characterized in that: The system has a program module corresponding to the steps of any one of the claims 1-7 above, and executes the steps in the above-described model predictive control-based unmanned aircraft trajectory planning and deployment decision method when running.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program configured to, when invoked by a processor, implement the steps of the unmanned aerial vehicle trajectory planning and deployment decision method based on model predictive control as described in any one of claims 1-7.

Images