A method and system for aerial loading planning for a transport helicopter

Abstract

This application discloses a method and system for in-flight loading planning for transport helicopters, specifically belonging to the field of helicopter loading planning technology. The method first determines an initial loading scheme based on helicopter performance parameters, cargo information, mission requirements, and preset loading rules. Secondly, it optimizes the initial loading scheme using an improved particle swarm optimization algorithm. Finally, during mission execution, based on dynamic changes in the mission, it updates the loading scheme by matching and executing corresponding adjustment operations in real time through a rule engine. This improved particle swarm optimization algorithm effectively balances the algorithm's global exploration capability with local convergence accuracy, avoiding getting trapped in local optima and significantly improving the optimization quality and convergence speed of the loading scheme. The introduction of a rule engine enables the system to respond to mission changes within seconds, achieving dynamic and local replanning of the loading scheme, thus improving the success rate and flexibility of mission execution.

Description

Technical Field

[0001] This application relates to the field of helicopter loading planning technology, and in particular to a method and system for aerial loading planning of transport helicopters. Background Technology

[0002] When transport helicopters perform missions such as cargo delivery, emergency rescue, and military transport, they need to rationally arrange cargo within limited cabin space and strict center of gravity constraints to ensure flight safety and mission efficiency. Traditional loading planning relies heavily on manual experience or simple rules, which has the following drawbacks: manual planning is time-consuming and difficult to adapt to complex scenarios involving multiple batches and types of cargo; it cannot respond to mission changes in real time and has poor dynamic adjustment capabilities; and it has low precision in controlling the helicopter's center of gravity, which can easily lead to decreased flight stability and even safety accidents.

[0003] In related technologies, heuristic algorithms such as the standard particle swarm optimization (PSO) are used for loading optimization. However, these algorithms are prone to getting trapped in local optima and have slow convergence speed, making it difficult to meet the requirements of real-time planning and dynamic adjustment in the air for algorithm performance and response time.

[0004] Therefore, there is an urgent need for an intelligent loading planning scheme that can take into account global optimization capabilities, rapid convergence characteristics, and dynamic adaptation capabilities to solve the above-mentioned technical problems. Summary of the Invention

[0005] In view of this, this application provides a method and system for aerial loading planning of transport helicopters, which can improve the optimization quality of loading schemes, planning efficiency, and dynamic response capability to mission changes.

[0006] Specifically, the following technical solutions are included:

[0007] This application provides a method for aerial loading planning of transport helicopters, the method comprising:

[0008] The initial loading plan is determined based on the helicopter's performance parameters, cargo information, mission requirements, and preset loading rules.

[0009] The initial loading scheme is optimized using an improved particle swarm optimization algorithm;

[0010] During task execution, the loading scheme is updated by matching and executing corresponding adjustment operations in real time based on dynamic changes in the task through the rule engine.

[0011] In some embodiments, a greedy algorithm is used to determine the initial loading scheme.

[0012] In some embodiments, the improved particle swarm optimization algorithm introduces dynamic inertia weights and adaptive learning factors during the iterative optimization process, and adopts parallel computing and elite retention strategies to output the optimal loading scheme.

[0013] In some embodiments, the iterative optimization process includes:

[0014] Set the population size, maximum number of iterations, particle dimension, and initialize particle position and velocity;

[0015] A multi-objective fitness function is designed based on multiple indicators, including load utilization rate, center of gravity deviation, and stability.

[0016] A nonlinear decreasing strategy is adopted to adjust the dynamic inertia weight;

[0017] Set the adaptive learning factor;

[0018] Parallel computation is performed using a master-slave parallel model. The problem is decomposed into multiple subproblems, and each subproblem is assigned a thread to evolve independently. The master process is used for population partitioning and result aggregation, while the slave processes are used to independently execute the subpopulation evolution. The problem is solved simultaneously through multi-threading, and an elite retention strategy is used to save the optimal solution in each iteration as the initial solution for subsequent iterations.

[0019] After each iteration, the global optimal solution is saved to the elite pool and the current worst individual is replaced; if the iteration termination condition is not met, the iterative optimization process is repeated.

[0020] Once the iteration termination condition is met, the optimal loading scheme is output after optimization.

[0021] In some embodiments, the dynamic inertia weight is adjusted according to the following formula:

[0022]

[0023] in, Represents dynamic inertia weight. Indicates the maximum inertia weight. This represents the minimum inertia weight. The adjustment coefficient is represented by t, the current iteration number is represented by t, and the maximum iteration number is represented by T.

[0024] Set the adaptive learning factor according to the following formula:

[0025]

[0026]

[0027] in, , These represent the first and second adaptive learning factors, respectively. , They represent , Initial value, , express , The final convergence value, where t represents the current iteration number and T represents the maximum number of iterations. , This represents the adjustment coefficient.

[0028] In some embodiments, the rule base of the rule engine includes a variety of dynamic event rules, including rules for adding emergency cargo loading, temporarily canceling cargo, loading emergency cargo, fuel consumption compensation, and center of gravity shift correction.

[0029] In some embodiments, in response to dynamic changes in the task, the rule engine matches the corresponding rule in the rule base, generates an adjustment instruction based on the corresponding rule, and triggers local optimization of the loading scheme.

[0030] In some embodiments, the helicopter performance parameters include the helicopter's maximum payload, cabin size, and center of gravity limit; the cargo information includes weight, volume, type, and priority; the mission requirements include destination, mileage, and time window; and the preset loading rules include dangerous goods isolation rules, large cargo priority placement rules, and center of gravity pre-allocation rules.

[0031] This application provides an aerial loading planning system for transport helicopters, the system comprising:

[0032] The loading plan generation module is used to determine the initial loading plan based on helicopter performance parameters, cargo information, mission requirements, and preset loading rules.

[0033] A convergence performance optimization module is used to optimize the initial loading scheme using an improved particle swarm optimization algorithm.

[0034] The dynamic adaptation module is used to update the loading scheme by matching and executing corresponding adjustment operations in real time according to the dynamic changes of the task during task execution through the rule engine.

[0035] The beneficial effects of the technical solutions provided in this application include at least the following:

[0036] This application provides a method and system for in-flight loading planning for transport helicopters. The method first determines an initial loading scheme based on helicopter performance parameters, cargo information, mission requirements, and preset loading rules. Second, it optimizes the initial loading scheme using an improved particle swarm optimization algorithm. Finally, during mission execution, based on dynamic mission changes, it updates the loading scheme by matching and executing corresponding adjustments in real time using a rule engine. The improved particle swarm optimization algorithm effectively balances the algorithm's global exploration capability with local convergence accuracy, avoiding getting trapped in local optima and significantly improving the optimization quality and convergence speed of the loading scheme. The introduction of a rule engine enables the system to respond to mission changes within seconds, achieving dynamic and local replanning of the loading scheme, greatly improving the success rate and flexibility of mission execution. The system's layered modular architecture decouples the functions of each module, facilitating maintenance, upgrades, and adaptation to different types of transport helicopters. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is a flowchart of a method for planning aerial loading of transport helicopters. Detailed Implementation

[0039] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0040] To make the technical solutions and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0041] This application provides a method for aerial loading planning of transport helicopters, such as... Figure 1 As shown, the method includes:

[0042] Step 101: Determine the initial loading plan based on the helicopter performance parameters, cargo information, mission requirements, and preset loading rules.

[0043] This step can quickly generate a feasible solution that satisfies the basic constraints, giving the algorithm a better starting point and avoiding ineffective searches. It provides high-quality elite guide individuals for subsequent improvements to the particle swarm optimization algorithm, which can shorten the early convergence time.

[0044] In some embodiments, helicopter performance parameters include the helicopter's maximum payload, cabin size, and center of gravity limit; cargo information includes weight, volume, type, and priority; mission requirements include destination, mileage, and time window; and preset loading rules include dangerous goods isolation rules, large cargo priority placement rules, and center of gravity pre-allocation rules.

[0045] Step 102: Optimize the initial loading scheme using an improved particle swarm optimization algorithm.

[0046] This step achieves an adaptive balance between global exploration and local convergence, effectively avoiding getting trapped in local optima and significantly improving the optimization quality of the loading scheme. The overall convergence speed is more than 30% faster than the standard particle swarm optimization algorithm.

[0047] In some embodiments, a greedy algorithm is used to determine the initial loading scheme.

[0048] In some embodiments, the improved particle swarm optimization algorithm introduces dynamic inertia weights and adaptive learning factors during the iterative optimization process, and adopts parallel computing and elite retention strategies to output the optimal loading scheme.

[0049] In some embodiments, the iterative optimization process includes:

[0050] Set the population size, maximum number of iterations, particle dimension, and initialize particle position and velocity;

[0051] A multi-objective fitness function is designed based on multiple indicators, including load utilization rate, center of gravity deviation, and stability.

[0052] A nonlinear decreasing strategy is adopted to adjust the dynamic inertia weight;

[0053] Set the adaptive learning factor;

[0054] Parallel computation is performed using a master-slave parallel model. The problem is decomposed into multiple subproblems, and each subproblem is assigned a thread to evolve independently. The master process is used for population partitioning and result aggregation, while the slave processes are used to independently execute the subpopulation evolution. The problem is solved simultaneously through multi-threading, and an elite retention strategy is used to save the optimal solution in each iteration as the initial solution for subsequent iterations.

[0055] After each iteration, the global optimal solution is saved to the elite pool and the current worst individual is replaced; if the iteration termination condition is not met, the iterative optimization process is repeated.

[0056] Once the iteration termination condition is met, the optimal loading scheme is output after optimization.

[0057] The algorithm enhances global search capability and convergence speed by introducing dynamic inertia weights and adaptive learning factors. Parallel computing and multi-threaded concurrent execution of fitness evaluation and particle updates significantly shorten the iteration time. Periodic information exchange between subgroups maintains the population's co-evolutionary ability, avoiding local optima caused by isolated evolution of subgroups. This allows the algorithm to complete hundreds of iterations within seconds, meeting the real-time planning requirements of the task. An elite retention strategy preserves the optimal solution in each iteration, ensuring that the optimal solution in each generation is never lost or degraded, guaranteeing monotonically convergent results (subsequent solutions are not inferior to previous ones). This effectively prevents the destruction of superior genes due to crossover, mutation, or velocity update operations, improving the algorithm's stability and the quality of the final solution.

[0058] In some embodiments, the dynamic inertia weight is adjusted according to the following formula:

[0059]

[0060] in, Represents dynamic inertia weight. Indicates the maximum inertia weight. This represents the minimum inertia weight. Let represent the adjustment coefficient, t represent the current iteration number, and T represent the maximum iteration number. (Nonlinear adjustment coefficient) The weight decay curve can be flexibly controlled to adapt to the optimization needs of problems of different scales.

[0061] In some embodiments, It can be 0.5. It can be 0.9. It can be 0.4.

[0062] In some embodiments, the particle dimension corresponds to the quantity of goods.

[0063] Set the adaptive learning factor according to the following formula:

[0064]

[0065]

[0066] in, , These represent the first and second adaptive learning factors, respectively. , They represent , Initial value, , express , The final convergence value, where t represents the current iteration number and T represents the maximum number of iterations. , This represents the adjustment coefficient.

[0067] Early stage > Particles rely more on their own historical best (individual cognition) to maintain population diversity and prevent premature maturation. Later... < Particles rely more on the group's historical best (social cognition), accelerating their aggregation towards the global optimum. This adaptive balance avoids the blind search or premature convergence caused by fixed learning factors.

[0068] Step 103: During task execution, based on dynamic changes in the task, the rule engine matches and executes corresponding adjustment operations in real time to update the loading scheme.

[0069] This step enables dynamic responses within seconds, without interrupting tasks for global replanning, thus improving task execution success rate and flexibility. The rule engine allows for flexible expansion of processing logic for different events, adapting to diverse real-world or emergency scenarios.

[0070] In some embodiments, the rule base of the rule engine includes a variety of dynamic event rules, including rules for adding emergency cargo loading, temporarily canceling cargo, emergency cargo loading, fuel consumption compensation, and center of gravity shift correction.

[0071] In some embodiments, in response to dynamic changes in the task, the rule engine matches the corresponding rule in the rule base, generates adjustment instructions based on the corresponding rule, and triggers local optimization of the loading scheme. Compared with global replanning, this significantly reduces the amount of computation, compresses the dynamic response time to the second level, and maintains the continuity and stability of the scheme.

[0072] In summary, the aerial loading planning method for transport helicopters provided in this application effectively balances the global exploration capability and local convergence accuracy of the particle swarm optimization algorithm by improving the algorithm, avoiding getting trapped in local optima, and significantly improving the optimization quality and convergence speed of the loading scheme. The introduction of a rule-based engine enables the system to respond to task changes in seconds, realize dynamic and local replanning of the loading scheme, and greatly improve the success rate and flexibility of task execution.

[0073] This application provides an aerial loading planning system for transport helicopters, the system comprising:

[0074] The loading plan generation module is used to determine the initial loading plan based on helicopter performance parameters, cargo information, mission requirements, and preset loading rules.

[0075] A convergence performance optimization module is used to optimize the initial loading scheme using an improved particle swarm optimization algorithm.

[0076] The dynamic adaptation module is used to update the loading scheme by matching and executing corresponding adjustment operations in real time according to the dynamic changes of the task during task execution through the rule engine.

[0077] In summary, the aerial loading planning system for transport helicopters provided in this application effectively balances the global exploration capability and local convergence accuracy of the particle swarm optimization algorithm by improving it, avoiding getting trapped in local optima and significantly improving the optimization quality and convergence speed of the loading scheme. The introduction of a rule-based engine enables the system to respond to task changes within seconds, achieving dynamic and local replanning of the loading scheme, thus improving the success rate and flexibility of task execution. The system's layered modular architecture decouples the functions of each module, facilitating maintenance, upgrades, and adaptation to different types of transport helicopters.

[0078] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the application disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only.

[0079] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A method for planning aerial loading of transport helicopters, characterized in that, The method includes: The initial loading plan is determined based on the helicopter's performance parameters, cargo information, mission requirements, and preset loading rules. The initial loading scheme is optimized using an improved particle swarm optimization algorithm; During task execution, the loading scheme is updated by matching and executing corresponding adjustment operations in real time based on dynamic changes in the task through the rule engine.

2. The aerial loading planning method for transport helicopters according to claim 1, characterized in that, A greedy algorithm is used to determine the initial loading scheme.

3. The aerial loading planning method for transport helicopters according to claim 1, characterized in that, The improved particle swarm optimization algorithm introduces dynamic inertia weights and adaptive learning factors during the iterative optimization process, and adopts parallel computing and elite retention strategies to output the optimal loading scheme.

4. The aerial loading planning method for transport helicopters according to claim 3, characterized in that, The iterative optimization process includes: Set the population size, maximum number of iterations, particle dimension, and initialize particle position and velocity; A multi-objective fitness function is designed based on multiple indicators, including load utilization rate, center of gravity deviation, and stability. A nonlinear decreasing strategy is adopted to adjust the dynamic inertia weight; Set the adaptive learning factor; Parallel computation is performed using a master-slave parallel model. The problem is decomposed into multiple subproblems, and each subproblem is assigned a thread to evolve independently. The master process is used for population partitioning and result aggregation, while the slave processes are used to independently execute the subpopulation evolution. The problem is solved simultaneously through multi-threading, and an elite retention strategy is used to save the optimal solution in each iteration as the initial solution for subsequent iterations. After each iteration, the global optimal solution is saved to the elite pool and the current worst individual is replaced; if the iteration termination condition is not met, the iterative optimization process is repeated. Once the iteration termination condition is met, the optimal loading scheme is output after optimization.

5. The aerial loading planning method for transport helicopters according to claim 4, characterized in that, Adjust the dynamic inertia weight according to the following formula: in, Represents dynamic inertia weight. Indicates the maximum inertia weight. This represents the minimum inertia weight. The adjustment coefficient is represented by t, the current iteration number is represented by t, and the maximum iteration number is represented by T. Set the adaptive learning factor according to the following formula: in, , These represent the first and second adaptive learning factors, respectively. , They represent , Initial value, , express , The final convergence value, where t represents the current iteration number and T represents the maximum number of iterations. , This represents the adjustment coefficient.

6. The aerial loading planning method for transport helicopters according to claim 1, characterized in that, The rule engine's rule base includes various dynamic event rules, including rules for adding emergency cargo loading, temporarily canceling cargo, loading emergency cargo, fuel consumption compensation, and center of gravity shift correction.

7. The aerial loading planning method for a transport helicopter according to claim 6, characterized in that, When dynamic changes occur in the task, the rule engine matches the corresponding rule in the rule base, generates adjustment instructions based on the corresponding rule, and triggers local optimization of the loading scheme.

8. The aerial loading planning method for a transport helicopter according to claim 1, characterized in that, The helicopter performance parameters include the helicopter's maximum payload, cabin size, and center of gravity limit; the cargo information includes weight, volume, type, and priority; the mission requirements include destination, mileage, and time window; and the preset loading rules include dangerous goods isolation rules, large cargo priority placement rules, and center of gravity pre-allocation rules.

9. An aerial loading planning system for transport helicopters, characterized in that, The system includes: The loading plan generation module is used to determine the initial loading plan based on helicopter performance parameters, cargo information, mission requirements, and preset loading rules. A convergence performance optimization module is used to optimize the initial loading scheme using an improved particle swarm optimization algorithm. The dynamic adaptation module is used to update the loading scheme by matching and executing corresponding adjustment operations in real time according to the dynamic changes of the task during task execution through the rule engine.

Images