An aircraft impulse orbit control fuel optimization method, device, equipment and medium based on a particle swarm algorithm

Abstract

The application discloses a kind of based on particle swarm algorithm's aircraft pulse orbit control fuel optimization method, device, equipment and medium, method includes the following steps: considering aircraft pulse orbit control thrust characteristic, for the requirement that guidance instruction is converted to orbit control pulse thrust instruction needs by continuous signal conversion to discrete signal, first, switch threshold parameterization is converted to function, design index vector is characterized in parameter in guidance instruction conversion function;Residual fuel is used as optimization index, design particle fitness function, and propose real-time updating strategy;Randomly generate conversion parameter index vector, as the initial population of particle swarm algorithm, optimize index vector by iteration, and obtain optimal conversion parameter after iteration is completed.The aircraft pulse orbit control particle swarm fuel optimization algorithm is used in the application, and the fuel utilization rate is improved by iterative optimization of the conversion parameter of guidance instruction.The application has the characteristics of simple design method, fuel limited adaptation, etc.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of pulse orbit control vehicle guidance system design, in particular, relates to a vehicle pulse orbit control fuel optimization method, device, equipment and medium based on a particle swarm algorithm. BACKGROUND

[0002] Space vehicles use pulse thrust for attitude and orbit adjustment, and realize attitude adjustment and orbit guidance process through direct force control. However, for some miniaturized space vehicles, the fuel carried by them is very limited. How to realize the performance requirements of attitude and orbit control under the condition of limited fuel, and avoid the problem of system oscillation caused by frequent switching of pulse thrust, has important engineering and theoretical research significance.

[0003] As one of the effective solutions, the fuel optimization of the vehicle based on trajectory optimization is a commonly used method. Fuel consumption is usually taken as the optimization index, and trajectory optimization method is used to solve the problem to realize trajectory planning under the condition of optimal index. This kind of method usually focuses on continuous thrust optimization problem, and is not suitable for pulse type space vehicle orbit control fuel optimization problem.

[0004] Space vehicles use pulse engines for guidance and orbit control, usually with fuel limitation, and need to design appropriate guidance law and corresponding instruction conversion function to reduce fuel consumption while completing target rendezvous. SUMMARY

[0005] In view of the above technical problems, the application designs a vehicle pulse orbit control fuel optimization method, device, equipment and medium based on particle swarm algorithm, which is a combined parameter optimization algorithm, designed for continuous guidance instruction conversion process, and the conversion parameters are optimized to reduce fuel consumption, and finally the particle swarm optimization method is used to solve the optimal parameters.

[0006] The vehicle pulse orbit control fuel optimization method based on particle swarm algorithm of the application comprises the following steps:

[0007] S1: for the guidance instruction conversion function of pulse orbit control, the conversion function switch threshold is parameterized, and the index vector is designed to represent the parameters in the guidance instruction conversion function;

[0008] S2: based on the remaining fuel and target interception state of single task, the fitness function is designed as follows:

[0009] (6)

[0010] Wherein, is the number of vehicles to complete the flight task, is a reward coefficient, is a fixed value and greater than the sum of the initial mass of the fuel carried by the aircraft; is the remaining fuel mass of the aircraft, is an index vector;

[0011] S3: randomly generate an index vector as an initial population, then perform iteration of the particle swarm optimization algorithm until the iteration termination condition is met, and finally obtain the conversion parameter corresponding to the maximum fitness in the particle swarm space, that is, output as the optimal solution, the conversion parameter refers to the parameter obtained after the conversion function switch threshold is parameterized in step S1.

[0012] The application also provides an aircraft impulse orbit control fuel optimization device based on a particle swarm algorithm, comprising the following modules:

[0013] A parameterization module is configured to parameterize a conversion function switch threshold for a guidance instruction conversion function of impulse orbit control, and design an index vector to represent parameters in the guidance instruction conversion function;

[0014] A function design module is configured to design a fitness function based on the remaining fuel obtained in a single task and a target interception state as follows:

[0015] (6)

[0016] wherein, is the number of aircrafts to complete the flight task, is a reward coefficient, is a fixed value and greater than the sum of the initial mass of the fuel carried by the aircraft; is the remaining fuel mass of the aircraft, is an index vector;

[0017] An iteration module is configured to randomly generate an index vector as an initial population, then perform iteration of the particle swarm optimization algorithm until the iteration termination condition is met, and finally obtain the conversion parameter corresponding to the maximum fitness in the particle swarm space, that is, output as the optimal solution, the conversion parameter refers to the parameter obtained after the conversion function switch threshold is parameterized in the parameterization module.

[0018] The application also provides a computing device comprising at least one processor and a memory storing program instructions; when the program instructions are read and executed by the processor, the computing device performs an aircraft impulse orbit control fuel optimization method based on a particle swarm algorithm.

[0019] The application also provides a readable storage medium storing program instructions, when the program instructions are read and executed by a computing device, the computing device performs an aircraft impulse orbit control fuel optimization method based on a particle swarm algorithm.

[0020] The application has the following beneficial effects:

[0021] The present application is directed to the problem of impulse orbit control fuel optimization of space vehicles, while considering the problem of too high state space dimension and too large data volume, and a heuristic algorithm is adopted to search for the optimal parameters of impulse orbit control fuel. The method designed in the present application is separated from the design of the guidance law, so that for any output continuous command guidance law, the present method can be used in the command conversion process for optimization. The present method is aimed at optimizing the parameters in the continuous guidance command conversion process, and particle swarm optimization is adopted to solve the problem, so as to realize the impulse orbit control requirement under fuel limitation. According to the technical scheme of the present application, for different types of impulse orbit control thrust and guidance law forms, only the state and fitness function need to be redesigned, and the optimal parameters of fuel can be iteratively updated, and the present application has good applicability. The aircraft impulse orbit control fuel optimization method based on particle swarm algorithm of the present application can be used for different types of impulse orbit control aircraft and can optimize the conversion function parameters to reduce fuel consumption. The present application adopts the particle swarm optimization algorithm, and has the advantages of simple design method and low cost. BRIEF DESCRIPTION OF DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort.

[0023] Figure 1 The flowchart of the aircraft impulse orbit control fuel optimization method based on particle swarm algorithm of the present application is shown in the figure.

[0024] Figure 2 The hysteresis characteristic curve in the conversion function is shown in the figure.

[0025] Figure 3 The three-dimensional interception scene of multiple space vehicles to multiple targets is shown in the figure.

[0026] Figure 4 The simulation schematic diagram before optimization is shown in the figure.

[0027] Figure 5 The simulation schematic diagram after optimization is shown in the figure. DETAILED DESCRIPTION

[0028] The technical solutions in the embodiments of the present application will be described clearly and completely in the embodiments of the present application combined with the drawings. Obviously, the described embodiments are only some of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art belong to the scope of protection of the present application.

[0029] As Figure 1As shown, the steps of the vehicle pulse orbit control fuel optimization method based on particle swarm optimization algorithm of the present invention are as follows: First, based on the characteristics of vehicle orbit control pulse thrust and the form of guidance law, the function switch threshold for converting guidance commands into engine commands is parameterized, and a conversion parameter index vector is designed; then, based on the remaining fuel and target interception state obtained from a single mission, a fitness function is designed. This fitness function takes the particle state index vector as input and outputs the particle fitness; finally, particle swarm optimization is performed based on the designed index vector and fitness function. First, an initial population is randomly generated and the fitness is calculated. Then, iteration is performed to maximize the particle fitness. After the iteration, the optimal conversion parameters in the particle swarm state space can be obtained. From the above steps, it can be seen that the evaluation criteria of the conversion parameters mainly depend on the fitness function of the corresponding particles. This value is determined by simulation. The more remaining fuel, the greater the fitness, which is consistent with the optimization requirements. The specific implementation steps are as follows:

[0030] S1: For the guidance command conversion function of pulse orbit control, the switching threshold of the conversion function is parameterized, and an index vector is designed to characterize the parameters in the guidance command conversion function.

[0031] Guidance command conversion functions include longitudinal and lateral conversion functions, both of which are typical hysteresis processes, such as... Figure 2 As shown, its expression is as follows:

[0032] (1)

[0033] in, This refers to the engine's maximum thrust or the rail control's maximum thrust. for The guidance commands input to the conversion function at any time are generally periodic samples of continuous commands, which need to be further discretized through the conversion function. , and This represents the threshold for turning the hysteresis loop on or off, used to control the output. Specifically, the longitudinal transformation function is denoted as... The lateral transformation function is denoted as Both have their own independent parameters.

[0034] Figure 2 middle and The threshold for the positive command switch is a hysteresis loop. When the engine is not running and the input quantity increases from small to large, the switch only activates when... The output becomes At this time, the engine is injecting forward; when the engine is operating in the forward direction, it only injects forward when the input decreases to... The engine will only be shut off under the following conditions. The reverse switching threshold works similarly. Figure 2 In represents the maximum output value.

[0035] The parameter in the conversion function refers to the parameter obtained after the conversion function switch threshold is parameterized, and the design index vector characterizes the parameter, which includes:

[0036] (2)

[0037] wherein, are parameters in the longitudinal instruction conversion function, is a lateral conversion function parameter, is a longitudinal and lateral conversion function common parameter.

[0038] The relationship between each parameter of the index vector and the conversion function is as follows:

[0039] For the longitudinal conversion function , the parameters of each parameter of the index vector and satisfy the following relationship:

[0040] (3)

[0041] For the lateral conversion function , the parameters of each parameter of the index vector and satisfy the following relationship:

[0042] (4)

[0043] For the longitudinal and lateral conversion functions, the index vector parameter takes the same value, still denoted as .

[0044] In an embodiment, the engine is a pulse engine, assuming that the engine start instruction is received from time, and the engine stop instruction is received at time, the thrust variation characteristic is:

[0045] (5)

[0046] wherein, is the longitudinal and lateral thrust value of the engine; is the maximum thrust value of the engine in steady state operation; is the ascending and descending time.

[0047] S2: design the fitness function based on the remaining fuel obtained from a single task and the target interception state.

[0048] The fitness function can be designed as follows:

[0049] (6)

[0050] in, The number of aircraft required to complete the flight mission. The reward coefficient is a fixed value that is greater than the sum of the initial fuel masses carried by all aircraft. The remaining fuel mass of the aircraft, subscript Indicates the aircraft number, It is an index vector.

[0051] Specifically, the function obtains simulation result data corresponding to the input parameters by calling the simulation once, and then calculates its fitness. Before iteration, the simulation scenario and corresponding variables need to be set, and the guidance law and attitude control method also need to be determined. During iteration, the initial scenario, variables, and guidance law settings need to be kept consistent, with only the input transformation parameters differing.

[0052] This step may also include: calculating and updating the fitness function in real time through numerical simulation, as follows:

[0053] Given a simulation scenario and parameters for the aircraft and target, the scenario and parameters should remain consistent throughout the iteration process. A six-degree-of-freedom simulation model of the aircraft and target is constructed within the simulation scenario, where the aircraft intercepts the target. Proportional guidance is used by default for interception, but a custom guidance law can also be selected. The simulation stops when the aircraft successfully intercepts the target or determines that the interception has failed; fuel consumption is no longer calculated, and the success or failure of the interception is recorded. A simulation cycle is complete when all aircraft simulations have stopped. After the simulation is complete, the simulation results are exported and substituted into the fitness function to obtain the final value.

[0054] S3: Randomly generate an index vector as the initial population, then iterate using the particle swarm optimization algorithm until the iteration termination condition is met. Finally, obtain the transformation parameters corresponding to the maximum fitness in the particle swarm space, which are then output as the optimal solution. The transformation parameters refer to the parameters obtained after parameterizing the switching threshold of the transformation function in step S1.

[0055] The flowchart of the particle swarm optimization algorithm is shown below:

[0056]

[0057] The first step is initialization, which includes defining and assigning values ​​to the control variables of the iterative algorithm. For the total number of iterations, The particle swarm size is the initial total number of particles. These are the particle state update coefficients, determining the direction of particle state updates. Initialization also includes initializing the particle states, typically using random generation to obtain initial values. Additionally, the optimal state and optimal global particle state need to be defined for updates in subsequent loops.

[0058] The process then proceeds iteratively. In each iteration, the fitness of each particle is calculated according to formula (6), and the optimal state of the particle is updated based on the fitness value. After the update is complete, the global optimal state is updated based on the optimal state of the particles, and then the particle state update coefficients are updated. Then, the particle state and particle velocity are updated sequentially.

[0059] After all iterations are completed, the global optimal state of the particle is obtained, which is the optimal state value, and corresponds to the optimal combination of transformation parameters obtained by the optimization algorithm.

[0060] In summary, step S3 may include:

[0061] The parameters and variables of the particle swarm optimization algorithm are initialized, including the number of iterations, population size, upper and lower bounds for each dimension of the index vector, upper and lower bounds for particle iteration speed, weight parameters controlling the particle swarm optimization algorithm, and initialization of variables during the iteration process. The particle swarm is randomly initialized, and the fitness function is called to calculate the fitness of each particle.

[0062] In each iteration, the index vector is updated based on the particle velocity as follows:

[0063] (7)

[0064] in, For the first Particle states during step iteration This represents the particle velocity.

[0065] The particle velocity is then updated according to the following formula:

[0066] (8)

[0067] in, This is a weighting factor that controls the direction of particle updates; To increase the randomness of the iteration process and avoid getting trapped in local optima, the matrix is ​​randomly assigned. and These represent the global optimal index vector and the particle optimal index vector up to the current iteration, respectively, with subscripts... Indicates the particle number.

[0068] Performing iteration: calculating the fitness of each particle, then updating the particle optimal index vector according to the fitness, after updating, updating the global optimal index vector according to the particle optimal index vector, then updating the weight factor and then updating the particle state and particle velocity in turn.

[0069] After completing all iterations, the particle global optimal index vector is obtained, which corresponds to the optimal conversion parameter.

[0070] In the above iteration, the iteration termination condition can be that the preset number of iterations is reached.

[0071] Further, when different task requirements are required, only the simulation scene needs to be adjusted, and the particle swarm optimization will be optimized according to the simulation results, so it can be applied to different scenes.

[0072] Further, when the number of targets is greater than the number of aircraft, the aircraft will select targets for interception according to the established rules, and then apply the scene where the number of targets is the same as the number of aircraft.

[0073] Also provided is an aircraft pulse orbit control fuel optimization device based on a particle swarm algorithm, comprising the following modules:

[0074] The parameterization module is used for parameterizing the conversion function switch threshold of the pulse orbit control guidance instruction conversion function, and the index vector is used to represent the parameters in the guidance instruction conversion function.

[0075] The function design module is used to design the fitness function based on the remaining fuel and target interception state of a single task as follows:

[0076] (6)

[0077] wherein, is the number of aircrafts to complete the flight task, is a reward coefficient, which is a fixed value and greater than the sum of the initial mass of the fuel carried by the aircraft; is the remaining fuel mass of the aircraft, is the index vector.

[0078] The iteration module is used to randomly generate an index vector as an initial population, and then perform iteration of the particle swarm optimization algorithm until the iteration termination condition is met, and finally obtain the conversion parameter corresponding to the maximum fitness in the particle swarm space, which is output as the optimal solution. The conversion parameter refers to the parameter obtained after parameterizing the conversion function switch threshold in the parameterization module.

[0079] A computing device comprises at least one processor and a memory storing program instructions; when the program instructions are read and executed by the processor, the computing device performs an aircraft pulse orbit control fuel optimization method based on a particle swarm algorithm.

[0080] A readable storage medium storing program instructions, when the program instructions are read and executed by a computing device, the computing device performs an aircraft pulse orbit control fuel optimization method based on a particle swarm algorithm.

[0081] The effectiveness of the method is described below with the example of a certain aircraft intercepting a non-maneuverable missile. The maximum thrust generated by the orbit control pulse engine is 3000N, the scene setting is shown in Table 1, and the simulation parameter setting is shown in Table 2. There are 6 aircrafts in the simulation setting, each intercepting an independent target. The end guidance process scene is shown in Table 1. Figure 3

[0082] Table 1 Simulation scene setting

[0083]

[0084] In Table 1, km and m, s are the units of distance, time, respectively.

[0085] Table 2 Simulation parameter setting

[0086]

[0087] In Table 2, kg, m, s are the units of mass, distance, time, respectively, and respectively represent the minimum and maximum values of each component of the index vector, and respectively represent the minimum and maximum values of the particle update speed component.

[0088] After the particle swarm optimization of the above parameters, the conversion parameters obtained by simulation can obtain the engine output thrust of the 6 aircrafts. Figure 4 Without parameter optimization, the simulation results of the pulse orbit control thrust curve are shown in Table 1, and the simulation curve after parameter optimization is shown in Table 1. Figure 5 The Fy / N in the table represents the longitudinal thrust of the aircraft, and the Fz / N represents the lateral thrust of the aircraft. By comparing the two, it can be seen that after optimization, the switching frequency of the engine is reduced.

[0089] The above only describes the preferred embodiments of the present application and does not limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.​

Claims

1. A method for optimizing fuel allocation in pulsed orbit control of an aircraft based on particle swarm optimization, characterized in that, Includes the following steps: S1: For the guidance command conversion function of pulse track control, the switching threshold of the conversion function is parameterized, and an index vector is designed to characterize the parameters in the guidance command conversion function; S2: Design a fitness function based on the remaining fuel and target interception status obtained from a single mission; S3: Randomly generate an index vector as the initial population, and then iterate the particle swarm optimization algorithm until the iteration termination condition is met. Finally, the transformation parameter corresponding to the maximum fitness in the particle swarm space is obtained, which is the output as the optimal solution. The transformation parameter refers to the parameter obtained after parameterizing the switching threshold of the transformation function in step S1. The fitness function is designed as follows: （6） in, The number of aircraft required to complete the flight mission. This is the reward coefficient, a fixed value that is greater than the sum of the initial mass of fuel carried by the aircraft; The remaining fuel mass of the aircraft It is an indicator vector; In step S1, the design index vector characterizes the parameters in the guidance command conversion function, including: （2） in, These are the parameters in the vertical instruction conversion function. For the lateral transformation function parameters, The parameters are shared by the longitudinal and lateral transformation functions; In step S1, both the longitudinal and lateral transformation functions are hysteresis processes: （1） in, This is the maximum thrust for track control; for The guidance command for the conversion function is input at all times; , and The hysteresis threshold is used to turn the hysteresis on or off, and the longitudinal transformation function is denoted as . The lateral transformation function is denoted as ; In step S1, the relationship between each parameter of the index vector and the transformation function is as follows: For vertical transformation function Indicator vector parameters and The parameters satisfy the following relationship: （3） For lateral transformation function Indicator vector parameters and The parameters satisfy the following relationship: （4） For vertical transformation functions and lateral transformation functions, the index vector parameters Values ​​and parameters of the conversion function If the values ​​are the same, they are still recorded as .

2. The spacecraft pulse orbit control fuel optimization method based on particle swarm optimization algorithm as described in claim 1, characterized in that, In step S2, the fitness function is calculated and updated in real time through numerical simulation. The numerical simulation process includes: Given a simulation scenario and parameters for the aircraft and target, a six-degree-of-freedom simulation model of the aircraft and target is constructed in the simulation scenario. The aircraft intercepts the target. When the aircraft successfully intercepts the target or determines that the interception has failed, the simulation of the aircraft stops and the fuel consumption is no longer calculated. At the same time, the success of the interception is recorded. When all aircraft simulations have stopped, a simulation is completed. After the simulation is completed, the simulation result data is exported and substituted into the fitness function to obtain the result value.

3. The spacecraft pulse orbit control fuel optimization method based on particle swarm optimization algorithm as described in claim 1, characterized in that, In step S3, the particle swarm algorithm steps are as follows: The parameters and variables in the particle swarm optimization algorithm are initialized, the particle swarm is randomly initialized, and the fitness function is called to calculate the fitness of each particle. In each iteration, the index vector is updated based on the particle velocity as follows: （7） in, For the first Particle states during step iteration The particle velocity; The particle velocity is then updated according to the following formula: （8） in, This is a weighting factor that controls the direction of particle updates; To increase the randomness of the iteration process and avoid getting trapped in local optima, the matrix is ​​randomly assigned. and These represent the global optimal index vector and the particle optimal index vector up to the current iteration round, respectively. Iteration: Calculate the fitness of each particle, then update the optimal metric vector of each particle based on its fitness. After updating, update the global optimal metric vector based on the optimal metric vector of each particle, and then update the weight factors. Then, the particle state and particle velocity are updated sequentially. After the iteration is completed, the globally optimal index vector is obtained, which corresponds to the optimal transformation parameters.

4. A vehicle pulse orbit control fuel optimization device based on particle swarm optimization algorithm, characterized in that, Includes the following modules: The parameterization module is used to parameterize the switching threshold of the guidance command conversion function for pulse orbit control and to design index vectors to characterize the parameters in the guidance command conversion function. The function design module is used to design the fitness function based on the remaining fuel and target interception status obtained from a single mission, as follows: （6） in, The number of aircraft required to complete the flight mission. This is the reward coefficient, a fixed value that is greater than the sum of the initial mass of fuel carried by the aircraft; The remaining fuel mass of the aircraft It is an indicator vector; The iteration module is used to randomly generate an index vector as the initial population, and then iterates the particle swarm optimization algorithm until the iteration termination condition is met. Finally, the transformation parameter corresponding to the maximum fitness in the particle swarm space is obtained, which is the output as the optimal solution. The transformation parameter refers to the parameter obtained by parameterizing the threshold of the transformation function in the parameterization module. In the parameterization module, the design index vector represents the parameters in the guidance command conversion function, including: （2） in, These are the parameters in the vertical instruction conversion function. For the lateral transformation function parameters, The parameters are shared by the longitudinal and lateral transformation functions; In the parameterization module, both the longitudinal and lateral transformation functions are hysteresis processes: （1） in, This is the maximum thrust for track control; for The guidance command for the conversion function is input at all times; , and The hysteresis threshold is used to turn the hysteresis on or off, and the longitudinal transformation function is denoted as . The lateral transformation function is denoted as ; In the parameterization module, the relationship between each parameter of the index vector and the transformation function is as follows: For vertical transformation function Indicator vector parameters and The parameters satisfy the following relationship: （3） For lateral transformation function Indicator vector parameters and The parameters satisfy the following relationship: （4） For vertical transformation functions and lateral transformation functions, the index vector parameters Values ​​and parameters of the conversion function If the values ​​are the same, they are still recorded as .

5. A computing device, characterized in that, include: At least one processor and a memory storing program instructions; When the program instructions are read and executed by the processor, the computing device performs the spacecraft pulse orbit control fuel optimization method based on the particle swarm algorithm as described in any one of claims 1-3.

6. A readable storage medium storing program instructions, characterized in that, When the program instructions are read and executed by the computing device, the computing device performs the spacecraft pulse orbit control fuel optimization method based on the particle swarm algorithm as described in any one of claims 1-3.

Images