A control method of an aircraft and related products

Abstract

The application discloses a control method of an aircraft and related products, and the method comprises the following steps: acquiring task demand information, operation parameter information and basic information of the aircraft in real time; the basic information comprises orbit position information, aircraft type information and equipment state information; comparing the basic information with a scheme template library to obtain a target operation control plan template; calculating and filling elements in the target operation control plan template based on the task demand information and the operation parameter information to obtain a plurality of alternative operation control plans; determining an operation control scheme based on the plurality of alternative operation control plans by using a multi-objective optimization algorithm; and controlling the aircraft based on the operation control scheme, which not only greatly reduces the dependence on artificial experience, but also realizes the automatic and intelligent generation of the operation control scheme, significantly improves the operation control scheme generation efficiency, response real-time performance and operation reliability, and enhances the autonomous operation capability of the aircraft in a high-dynamic and strong-constraint space environment.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of aircraft, and in particular to a control method of an aircraft and related products. BACKGROUND

[0002] In the existing aircraft operation control scheme generation technology, artificial experience or static preset rules are generally relied on for preplan design, and there are problems such as low automation, poor adaptability and insufficient optimization capability. For example, it is usually necessary for an operator to manually convert a task instruction described in natural language into structured information, which not only consumes time and effort, but also may easily lead to missing of key elements, semantic misjudgment or incomplete information due to individual understanding differences of the operator, and thus cause the generated preplan to deviate from the actual task requirements. Simple matching mechanisms based on fixed templates are mostly adopted, and only a small number of discrete parameters such as task types and aircraft models are matched in a coarse-grained manner, which is difficult to cope with complex and variable task scenarios and dynamically changing aircraft states. Engineers are generally relied on to manually adjust parameters based on experience, or single traditional optimization algorithms are used for local calculation, and there is a lack of comprehensive consideration of multi-dimensional factors such as task objectives, real-time aircraft states, orbital environments and energy constraints, and it is difficult to achieve global collaborative optimization of operation control parameters.

[0003] To sum up, the existing technology highly relies on manual intervention, lacks automatic analysis and intelligent decision-making capability for task and state information, and leads to low efficiency, slow response speed and weak adaptability of operation control scheme generation, which is difficult to meet the autonomous and intelligent operation control requirements of low-orbit aircraft in a high-dynamic, multi-task and strong real-time environment. SUMMARY

[0004] Based on the above problems, the present application provides a control method of an aircraft and related products, aiming to improve the generation efficiency of aircraft operation control.

[0005] The embodiments of the present application disclose the following technical solutions:

[0006] The first aspect of the present application provides a control method of an aircraft, which comprises:

[0007] real-time acquisition of task requirement information, running parameter information and basic information of the aircraft; the basic information comprises orbital position information, aircraft type information and equipment state information;

[0008] comparing the basic information with a scheme template library to obtain a target operation control preplan template;

[0009] calculating and filling elements in the target operation control preplan template based on the task requirement information and the running parameter information to obtain a plurality of alternative operation control preplans;

[0010] determining a control scheme based on the multiple candidate control plans by using a multi-objective optimization algorithm;

[0011] controlling the aircraft based on the control scheme.

[0012] Optionally, the comparing the basic information and the scheme template library obtains a target control plan template, and specifically includes:

[0013] performing vector conversion on the basic information to obtain a feature vector; the feature vector includes a feature vector corresponding to the basic information;

[0014] performing normalization and standardization processing on the feature vector to obtain a processed feature vector;

[0015] dynamically calculating an association weight between the processed feature vectors based on an attention mechanism to obtain a key vector;

[0016] comparing the key vector and the scheme template library based on a similarity matching algorithm to obtain a target control plan template; the scheme template library includes control plan templates of various aircrafts in different task scenarios.

[0017] Optionally, the calculating and filling elements in the target control plan template based on the task requirement information and the operation parameter information obtains multiple candidate control plans, and specifically includes:

[0018] performing text segmentation on the task requirement information to obtain multiple sub-information; the sub-information includes a single word or phrase;

[0019] mapping each sub-information to obtain a vector corresponding to each sub-information;

[0020] analyzing and determining semantic associations between the sub-information based on the vector corresponding to each sub-information to obtain key elements corresponding to the task requirement information;

[0021] inputting the key elements and the operation parameter information into an element model to obtain multiple sets of to-be-supplemented elements;

[0022] obtaining the multiple candidate control plans based on the multiple sets of to-be-supplemented elements and the target control plan template; the number of the sets of to-be-supplemented elements is consistent with the number of the candidate control plans.

[0023] Optionally, the method further includes:

[0024] obtaining control plan templates of various aircrafts in different task scenarios;

[0025] Analyze the operation and control plan templates of the various types of aircraft in different task scenarios to determine common operations and parameter settings in the operation and control plan templates of the various types of aircraft in different task scenarios;

[0026] Abstract the common operations and the parameter settings as general data;

[0027] Separate the common operations and parameter settings corresponding to the general data from the operation and control plan templates of the various types of aircraft in different task scenarios to obtain the separated operation and control plan templates of the various types of aircraft in different task scenarios;

[0028] Classify the separated operation and control plan templates of the various types of aircraft in different task scenarios and label the classification information in the separated operation and control plan templates of the various types of aircraft in different task scenarios;

[0029] Construct a scheme template library based on the labeled operation and control plan templates of the various types of aircraft in different task scenarios.

[0030] Optionally, the operation and control scheme is determined based on the multiple candidate operation and control plans by using a multi-objective optimization algorithm, specifically including:

[0031] Constructing a multi-objective function based on the multiple candidate operation and control plans;

[0032] Constructing a constraint condition;

[0033] Obtaining a task priority and adjusting the weight of the multi-objective function based on the task priority and the basic information to obtain an adjusted multi-objective function;

[0034] Solving the adjusted multi-objective function based on the constraint condition by using an improved non-dominated sorting genetic algorithm to determine the operation and control scheme.

[0035] Optionally, the operation and control scheme is determined based on the constraint condition by solving the adjusted multi-objective function using an improved non-dominated sorting genetic algorithm, specifically including:

[0036] Let the value of k be 1;

[0037] Construct an initial population based on the adjusted multi-objective function, and use the initial population as the population of the kth iteration; the initial population includes multiple individuals; the multiple individuals represent multiple solutions of the adjusted multi-objective function; the multiple solutions represent multiple candidate operation and control plans;

[0038] Quickly non-dominantly sort the population of the kth iteration to obtain a non-dominantly sorted population of the kth iteration;

[0039] calculate the crowding distance of each individual in the non-dominated sorted population of the kth iteration, and select the non-dominated sorted population of the kth iteration based on the constraint condition and the crowding distance of each individual, to obtain a selected population of the kth iteration;

[0040] perform cross and mutation processing on the selected population of the kth iteration by adopting tournament selection, simulated binary crossover and polynomial mutation, to obtain a sub-population of the kth iteration;

[0041] perform screening on the sub-population of the kth iteration by adopting an elitist strategy, to obtain a final population of the kth iteration;

[0042] if the iteration end condition is met, output the individual in the final population of the kth iteration as the operation and control scheme;

[0043] if the iteration end condition is not met, take the final population of the kth iteration as the population of the (k+1)th iteration, let the value of k be 1, and return to the step of performing fast non-dominated sorting on the population of the kth iteration to obtain the non-dominated sorted population of the kth iteration.

[0044] The second aspect of the application provides a control device of a spacecraft, and the control device of the spacecraft comprises:

[0045] an acquisition module, configured to acquire task demand information, operation parameter information and basic information of the spacecraft in real time; the basic information comprises orbit position information, spacecraft type information and equipment state information;

[0046] a comparison module, configured to compare the basic information and a scheme template library to obtain a target operation and control plan template;

[0047] a filling module, configured to calculate and fill elements in the target operation and control plan template based on the task demand information and the operation parameter information, to obtain a plurality of alternative operation and control plans;

[0048] an operation and control scheme determination module, configured to determine an operation and control scheme based on the plurality of alternative operation and control plans by adopting a multi-objective optimization algorithm;

[0049] a control module, configured to control the spacecraft based on the operation and control scheme.

[0050] The third aspect of the application provides a computer device, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor executes the computer program to implement the control method of the spacecraft provided in the first aspect.

[0051] The fourth aspect of the application provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the control method of the aircraft provided in the first aspect.

[0052] The fifth aspect of the application provides a computer program product, which comprises a computer program, and the computer program is executed by a processor to realize the control method of the aircraft provided in the first aspect.

[0053] Compared with the prior art, the application has the following beneficial effects:

[0054] The application comprises obtaining task demand information, running parameter information and basic information of the aircraft in real time; the basic information comprises orbit position information, aircraft type information and equipment state information; the basic information and a scheme template library are compared to obtain a target operation control plan template; elements in the target operation control plan template are calculated and filled based on the task demand information and the running parameter information to obtain a plurality of alternative operation control plans; a multi-objective optimization algorithm is used to determine an operation control scheme based on the plurality of alternative operation control plans; and the aircraft is controlled based on the operation control scheme.

[0055] The application realizes comprehensive perception of the running state of the aircraft by obtaining task demand information, running parameter information and basic information (comprising orbit position information, aircraft type information and equipment state information) of the aircraft in real time. The application improves the accuracy and response efficiency of plan selection by intelligently matching the basic information with a scheme template library to quickly determine an adaptive target operation control plan template. On this basis, the application dynamically calculates and fills key elements in the plan template in combination with specific task demand and real-time running parameter to generate a plurality of alternative operation control plans that meet the current working condition, thereby enhancing the pertinence and feasibility of the scheme. Further, the application uses a multi-objective optimization algorithm to comprehensively evaluate the alternative operation control plans and determine an optimal operation control scheme. Finally, the aircraft is accurately controlled based on the scheme, which significantly improves the operation control scheme generation efficiency, response real-time performance and running reliability, enhances the autonomous running capability of the aircraft in a high dynamic and strong constraint space environment, and has outstanding substantial features and significant technical progress. BRIEF DESCRIPTION OF DRAWINGS

[0056] In order to more clearly illustrate the technical solutions in the embodiments of the application or the prior art, the drawings needed in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative labor.

[0057] Figure 1A flow chart of a control method of a spacecraft provided for an embodiment of the present application;

[0058] Figure 2 A structural diagram of a control device of a spacecraft provided for an embodiment of the present application. DETAILED DESCRIPTION

[0059] As described above, the existing spacecraft operation control scheme generation technology mainly relies on artificial experience or fixed rules, has low automation degree, poor adaptability and weak optimization capability. Task analysis mainly relies on artificial conversion of natural language instructions into structured information, which is low in efficiency and prone to understanding deviation and element omission, resulting in disconnection between the operation control scheme and actual demand; operation control scheme matching generally adopts template matching based on a small number of parameters such as task type and spacecraft model, which is rigid in mode and difficult to cope with complex dynamic tasks and flight state changes; parameter adjustment relies on artificial experience or single algorithm, lacks comprehensive optimization of multiple factors such as task target, real-time state, environment and energy, and is difficult to achieve global collaboration. The overall technology highly depends on artificial intervention, lacks intelligent analysis and decision-making capability, resulting in low efficiency, slow response and insufficient adaptability of the operation control scheme generation, which is difficult to meet the autonomous and intelligent operation control needs of the low-orbit spacecraft in the high-dynamic, multi-task and strong real-time scenarios.

[0060] In view of the above problems, the present application provides a control generation method of a spacecraft and related products, the method comprising: acquiring task demand information, running parameter information and basic information of the spacecraft in real time; the basic information comprises orbit position information, spacecraft type information and equipment state information; comparing the basic information with a scheme template library to obtain a target operation control plan template; calculating and filling elements in the target operation control plan template based on the task demand information and the running parameter information to obtain a plurality of alternative operation control plans; determining an operation control scheme based on the plurality of alternative operation control plans by using a multi-objective optimization algorithm; and controlling the spacecraft based on the operation control scheme.

[0061] The application can automatically analyze and process task demand information, intelligently identify and extract core key elements affecting operation control decision, realize efficient conversion of unstructured task instructions into structured decision input, and improve the accuracy and efficiency of information processing. In the operation control scheme generation link, the traditional static matching mode relying on fixed templates is abandoned, and the target operation control plan template is quickly determined through intelligent matching with the scheme template library, and combined with task demand and operation parameter information for dynamic filling to generate multiple alternative operation control plans. A multi-objective optimization algorithm is introduced, and multiple targets such as flight efficiency, energy consumption, flight safety and task completion quality are considered to globally optimize the multiple alternative operation control plans and select the operation control scheme. Finally, based on the operation control scheme, precise control is implemented on the aircraft, realizing the whole process cooperation from task understanding, state perception, intelligent decision to closed-loop control. The dependence on artificial experience is effectively reduced, the automatic and intelligent generation of operation control plan is realized, the generation efficiency, response speed and operation reliability are greatly improved, and the autonomous decision and task execution capability of the aircraft in high dynamic and strong constraint space environment is significantly enhanced.

[0062] In order to enable those skilled in the art to better understand the scheme of the present application, the technical solutions in the embodiments of the present application will be described clearly and completely in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part 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 without creative labor are within the scope of protection of the present application.

[0063] Figure 1 A flow chart of a control method of a spacecraft provided by the embodiments of the present application is shown in Figure 1 The control method of the spacecraft includes:

[0064] S101: Real-time acquisition of task demand information, operation parameter information and basic information of the spacecraft.

[0065] The type of the spacecraft is not limited by the present application, for example, the spacecraft is a low-orbit spacecraft; the acquisition method of the task demand information, operation parameter information and basic information is not limited by the present application, for example, the original task instruction (such as natural language or structured text) of the spacecraft is received through a ground command system or a satellite-borne task management system, and the operation parameter information and basic information of the spacecraft are collected in real time through a satellite-ground link or an on-board sensor. The basic information generally refers to some flight state data of the spacecraft at present, for example, the basic information includes orbit position information (latitude, longitude, height, etc.), aircraft type information (fixed-wing unmanned aerial vehicle, rotary-wing unmanned aerial vehicle, vertical take-off and landing unmanned aerial vehicle and missile, etc.) and equipment state information (energy storage, sensor working state, etc.).

[0066] The comprehensive dynamic perception of the flight task intention and the current running state is realized, accurate and timely data basis is provided for subsequent intelligent decision-making, and control deviation caused by information lag or loss is avoided.

[0067] S102: The basic information and the scheme template library are compared to obtain a target operation and control plan template.

[0068] The scheme template library includes various operation and control plan templates, and the target operation and control plan template can be determined based on the basic information.

[0069] S103: Based on the task demand information and the running parameter information, elements in the target operation and control plan template are calculated and filled to obtain a plurality of alternative operation and control plans.

[0070] Since the target operation and control plan template may lack key data, the elements in the target operation and control plan template need to be filled based on the task demand information and the running parameter information, so as to determine a plurality of alternative operation and control plans.

[0071] S104: An operation and control scheme is determined based on the plurality of alternative operation and control plans by using a multi-objective optimization algorithm.

[0072] There are various methods for determining the operation and control scheme by using the multi-objective optimization algorithm, for example, a multi-objective optimization model is constructed based on the plurality of alternative operation and control plans, a plurality of conflicting objective functions (such as a minimum fuel consumption objective function, a maximum task completion degree objective function, a minimum execution time objective function, and a maximum flight safety objective function (such as an obstacle avoidance distance and a thermal load margin)) are defined; an improved multi-objective evolutionary algorithm (such as NSGA-II, MOEA / D, and an adaptive genetic algorithm) is used to optimize the plurality of conflicting objective functions, the optimal solution is selected by a non-dominated sorting and an improved crowding degree mechanism, and the operation and control scheme is output to an operator in a visual and easy-to-understand manner. A weight adaptive mechanism can also be introduced, and the weights of the objectives are dynamically adjusted according to the task priority. Global collaborative optimization of the operation and control scheme is realized, resource waste or risk accumulation caused by single index optimization is avoided, and optimal task execution under multiple constraint conditions is ensured.

[0073] S105: The aircraft is controlled based on the operation and control scheme.

[0074] The control method is not limited in the present application, for example, the operation control scheme is integrated with the task demand information and the basic information to generate a complete low-orbit spacecraft operation control scheme, ensuring the completeness and logicality of the operation control scheme content. The operation control scheme includes detailed contents such as the maneuver plan of the spacecraft (such as flight position, speed and attitude control instructions at each time period), task execution steps (such as sensor start and stop time, data acquisition frequency and data transmission mode, etc.), device operation instructions (such as device power-on and power-off time and working mode switching instructions, etc.), and the complete low-orbit spacecraft operation control scheme is output to the operator in a visual and easy-to-understand manner, supporting the operator to view and modify the operation control scheme. Seamless connection from intelligent decision-making to precise execution is realized, which guarantees the effective landing of the operation control scheme; through the closed-loop feedback mechanism, the control reliability is enhanced and the task execution success rate is improved.

[0075] The present application realizes comprehensive perception of flight task intention and current running state by real-time acquisition of task demand information and basic information (including orbit position information, spacecraft type information and device state information, etc.) of the spacecraft. By comparing the basic information and the scheme template library, the target operation control plan template is obtained, which breaks through the limitation of traditional fixed template matching and significantly improves the accuracy and environmental adaptability of the operation control scheme generation. Further, a multi-objective optimization algorithm is used to comprehensively weigh multiple objectives such as flight efficiency, energy consumption, safety and task completion degree, and select the operation control scheme from multiple alternative operation control plans, realizing the collaborative optimization of operation control parameters. Finally, based on the operation control scheme, precise control is implemented on the spacecraft, forming a closed-loop system of "task-state-decision-control". The automatic and intelligent generation of the operation control scheme is realized, greatly reducing the dependence on human experience, and significantly improving the operation control scheme generation efficiency, response real-time performance and task execution reliability.

[0076] The above is the main technical scheme of the present application, and the expansion implementation of the main technical scheme of the present application will be introduced. Specifically as follows:

[0077] For S102, the present application provides an optional embodiment for comparing the basic information and the scheme template library to obtain a target operation control plan template:

[0078] The basic information is converted into a feature vector, and the feature vector includes a feature vector corresponding to the basic information.

[0079] The basic information is converted into a corresponding feature vector by a large model.

[0080] The feature vector is normalized and standardized to obtain a processed feature vector.

[0081] Eliminate the weight deviation caused by the difference in dimension and order of magnitude between different features, avoid the dominance of certain features in the optimization or matching process; improve the accuracy and stability of the subsequent attention mechanism and similarity matching, enhance the model convergence and generalization ability.

[0082] Based on the attention mechanism, the correlation weight between the processed feature vectors is dynamically calculated to obtain the key vector.

[0083] Multi-Head Self-Attention architecture can be used to dynamically capture the correlation weight between different features to obtain the key vector. The mathematical expression is as follows:

[0084] MultiHead(Q,K,V)=Concat(head1,...,head h )W;

[0085] head i =Attention(QW i Q ,KW i K ,VW i V );

[0086] ;

[0087] Where MultiHead represents the multi-head self-attention mechanism; Q, K, and V are query matrix, key matrix, and value matrix respectively; Concat represents concatenation operation; head1 represents the output of the first self-attention head; head h represents the output of the hth self-attention head; head i represents the output of the ith self-attention head; W, W i Q , W i K , W i V are all learnable parameter matrices; Attention represents a single-head self-attention mechanism; d k is the key vector dimension; K T represents the transpose of K; softmax represents the softmax function.

[0088] The core role of the attention mechanism is to dynamically adjust the weight of each feature according to the task type and aircraft state. This dynamic weight adjustment mechanism enables the large model to focus on key features, improving the accuracy of plan matching and parameter adjustment.

[0089] The key vector is compared with the scheme template library based on a similarity matching algorithm to obtain a target operation control plan template; the scheme template library includes operation control plan templates of various aircrafts in different task scenarios.

[0090] The task demand information and the basic information are comprehensively considered and compared with the scheme template library by using a hybrid feature representation and a multi-modal similarity matching model to obtain the target operation control plan template. The specific design is as follows:

[0091] The feature representation is constructed, and for the feature vector T corresponding to the key element:

[0092] T=[t1,t2,...,t n ];

[0093] Wherein, t1 represents the task type, target area and time requirement and the like features of the first element in the feature vector corresponding to the key element; t2 represents the task type, target area and time requirement and the like features of the second element in the feature vector corresponding to the key element; t n n represents the task type, target area and time requirement and the like features of the n-th element in the feature vector corresponding to the key element.

[0094] For the feature vector A corresponding to the basic information:

[0095] ;

[0096] Wherein, a1 represents the position, attitude and device state and the like features of the first element in the feature vector corresponding to the basic information; a2 represents the position, attitude and device state and the like features of the second element in the feature vector corresponding to the basic information; n represents the position, attitude and device state and the like features of the n-th element in the feature vector corresponding to the basic information.

[0097] For the template feature vector M:

[0098] ;

[0099] Wherein, m1 represents the features of the operation control plan template of the first element in the template feature vector; m2 represents the features of the operation control plan template of the second element in the template feature vector; n represents the features of the operation control plan template of the n-th element in the template feature vector.

[0100] For the similarity matching algorithm, the following formula is used:

[0101] Sim(T,A,M)=α⋅Sim T (T,M T )+β⋅Sim​​A (A,M A )+γ⋅Sim TA (T,A,M);

[0102] wherein, Sim(T,A,M) represents the similarity calculated based on T, A and M; Sim T represents the similarity of the feature vectors corresponding to the key elements, calculated by cosine similarity; Sim A represents the similarity of the feature vectors corresponding to the basic information, calculated by the inverse of Euclidean distance; Sim TA represents the joint similarity of the feature vectors corresponding to the basic information and the key elements, calculated by semantic matching based on attention mechanism; M T represents the task feature vector of the operation control plan template, corresponding to the attributes related to the task in the operation control plan template (such as task type adaptation range and task target parameters, etc.), used for similarity calculation with T; M A represents the aircraft state feature vector of the operation control plan template, corresponding to the attributes related to the aircraft state in the operation control plan template (such as applicable position range and equipment state parameters, etc.), used for similarity calculation with A; α, β, γ are all weight coefficients, determined by historical data training.

[0103] In the similarity calculation, a hierarchical sorting strategy is also used, and the operation control plan templates with similarity greater than the threshold are preliminarily screened out based on the fast screening algorithm. Fine matching is performed on the operation control plan templates, and the comprehensive similarity score is calculated. According to the score, the first operation control plan template is selected as the target operation control plan template. This hybrid feature representation and hierarchical matching strategy can effectively improve the accuracy and efficiency of target operation control plan template matching. For an operation control task, according to the characteristics such as the coverage range, communication bandwidth requirement, and remote sensing action behavior of the task, the corresponding features of the corresponding task templates in the scheme template library are matched, and the template with the highest similarity is found. According to the similarity score, the top several scheme templates with the highest scores are automatically selected by the large model as the candidate operation control plan templates. The number of selected templates can be adjusted according to actual needs, providing multiple reference schemes for subsequent parameter adjustment.

[0104] The above describes in more detail the method of obtaining the target operation control plan template. Now the method of determining multiple alternative operation control plans based on the target operation control plan template obtained above will be described in detail, i.e., calculating and filling the elements in the target operation control plan template based on the task requirement information and the operation parameter information to obtain multiple alternative operation control plans. The present application provides an optional embodiment:

[0105] performing text segmentation on the task requirement information to obtain multiple sub-information; the sub-information includes single words or phrases.

[0106] The present application does not limit the text segmentation method, for example, using a large model word segmentation algorithm to segment the task requirement information into single words or phrases; or using a word segmentation technology (such as rule-based regular segmentation, Chinese word segmentation tool or English word segmentation tool) to finely segment the task requirement information to generate multiple sub-information.

[0107] Through fine-grained text segmentation, the unstructured task instruction is decomposed into a computable basic semantic unit, providing a basic input for subsequent vectorization and semantic analysis; avoids the analysis error caused by ambiguous sentence semantics or information coupling, and improves the accuracy and flexibility of task element extraction.

[0108] Each sub-information is mapped to obtain a vector corresponding to each sub-information.

[0109] For the mapping method, each sub-information can be mapped to a corresponding high-dimensional vector based on a pre-trained word vector model (such as a neural network model or a large language model), so that the text information is converted into a numerical form that can be processed by a computer; or for phrases or compound words (such as "high-resolution imaging"), an average pooling output phrase-level vector is used; or a context-aware model is used to encode each sub-information separately to obtain its context-independent but semantically rich high-dimensional representation (such as a 768-dimensional or 1024-dimensional vector). Discrete language symbols are converted into mathematical vectors in continuous space, realizing the computability of semantic information; high-dimensional vectors can capture the semantic features of words and phrases (such as synonym, synonym and category relationship), providing a mathematical basis for subsequent semantic correlation analysis, significantly improving the intelligent level of natural language understanding.

[0110] The vectors corresponding to each sub-information are analyzed to determine the semantic correlation between each sub-information, and the key elements corresponding to the task requirement information are obtained.

[0111] For the determination of key elements, a self-attention mechanism in a large model can be used to deeply analyze each sub-information to capture the semantic correlation between words, thereby identifying key elements such as task type, task target, and task time requirement. The language understanding ability of the large model is used to convert the task requirement described in natural language into structured task information, extract the key elements of the task, including task type (communication task, remote sensing monitoring task, and orbit maneuver task, etc.), task target (coverage area and monitoring accuracy, etc.), and task time requirement, and convert them into a structured data format for storage in the system for use by subsequent modules.

[0112] The key elements and the operation parameter information are input into an element model to obtain a plurality of sets of to-be-supplemented elements.

[0113] For example, the key elements include remote sensing monitoring tasks and covering the core area of the disaster area, spatial resolution ≤1.5m and first imaging completed within 30 minutes. The running parameter information (for example, a satellite) includes the current orbital position (altitude 550km, inclination 98°, expected to enter the target area within 12 minutes), attitude adjustment capability (can complete the pointing adjustment within 6 minutes) and energy state (battery power 85%).

[0114] The key elements and the running parameter information are input into an element model (a machine learning model for generating possible parameter combinations according to the input) to obtain a plurality of to-be-supplemented element sets (such as a first to-be-supplemented element set, a second to-be-supplemented element set and a third to-be-supplemented element set); the to-be-supplemented element set is not limited in the application, for example, the to-be-supplemented element set includes an imaging time window, a side-looking angle, an imaging mode, a data transmission strategy and an energy reservation.

[0115] Based on the plurality of to-be-supplemented element sets and the target operation control plan template, the plurality of alternative operation control plans are obtained.

[0116] Each to-be-supplemented element set is supplemented into the target operation control plan template to obtain a corresponding alternative operation control plan. In short, one to-be-supplemented element set corresponds to one alternative operation control plan, that is, the number of to-be-supplemented element sets and the number of alternative operation control plans are consistent.

[0117] By deeply analyzing the semantic association between high-dimensional vectors, the limitations of traditional keyword matching are broken through, the implicit logic and context dependence in complex sentences can be understood, and the key elements (such as targets, time, priority and resource requirements) affecting operation control decisions can be accurately extracted; the intelligent and automated level of task analysis is improved, manual intervention is reduced, and the adaptability of the system to diversified task instructions is enhanced.

[0118] The above embodiment mentions a scheme template library, and the scheme template library includes operation control plan templates of various aircrafts in different task scenarios. In order to ensure the effectiveness of the scheme template library, the application further provides an optional embodiment of constructing a scheme template library:

[0119] This embodiment mainly abstracts, optimizes and classifies the input historical operation control scheme cases to generate a corresponding scheme template library with detailed annotations and classification information.

[0120] Obtain operation control plan templates of various aircrafts in different task scenarios.

[0121] The operation control plan templates of various aircraft (such as satellites and unmanned aerial vehicles) in different task scenarios (such as imaging, communication and obstacle avoidance) can be collected through ground station telemetry, simulation system generation and task planning tool export, etc. Including but not limited to orbit adjustment instructions, attitude control sequences, payload configuration, energy management and communication plans, etc. The operation control plan templates are usually stored in a structured form (such as XML, JSON and CSV), including timestamp, operation type and parameter value fields.

[0122] By comprehensively collecting operation control plan templates of various aircraft in different task scenarios, a rich knowledge base is constructed; Ensure that the data source of subsequent analysis and template construction is extensive and representative, support intelligent matching and optimization of diversified task requirements.

[0123] The operation control plan templates of various aircraft in different task scenarios are analyzed to determine the common operations and parameter settings in the operation control plan templates of various aircraft in different task scenarios.

[0124] Common operations in the operation control plan templates are identified using statistical analysis, clustering analysis (such as K-means, hierarchical clustering) or association rule mining (such as Apriori algorithm). For example, for imaging tasks, common operations include pointing to target areas, adjusting camera resolution and setting data transmission priority. Parameter settings, such as power distribution strategy, fuel consumption threshold and attitude adjustment angle, etc. Extract high-frequency operation sequences and parameter configurations as common features. For example, the common parameter settings in the imaging task are "resolution = high" and "transmission bandwidth = maximum". Through in-depth analysis of a large number of operation control plan templates, common operations and parameter settings across tasks and platforms are identified and refined, providing a basis for subsequent abstract general templates; Reduce the amount of repetitive design work, improve the efficiency and consistency of operation control scheme generation.

[0125] The common operations and parameter settings are abstracted as general data.

[0126] The identified common operations and parameter settings are abstracted as general data. For example, create a general data task template that includes pointing to targets, adjusting cameras and setting transmission priorities. General data can be standardized operation sequences (such as JSON format task instructions), parameter configuration tables (such as XML format configuration files) or template definitions based on domain models (such as UML diagrams). Design for scalability and flexibility, allowing dynamic adjustment of some parameters (such as resolution and transmission bandwidth) according to different task requirements, simplifying the design process of complex tasks; Improve the reusability and maintainability of operation control schemes, reduce the probability of human error, enhance the robustness and adaptability of the system.

[0127] The commonality operation and parameter setting corresponding to the general data are abstracted from the operation and control plan templates of the various types of aircrafts in different task scenarios, to obtain the abstracted operation and control plan templates of the various types of aircrafts in different task scenarios.

[0128] The part that has been abstracted as general data is removed from the original operation and control plan template, and the operation and parameter setting specific to the task scenario are retained. For example, for the imaging task of a certain satellite, only the specific operation (such as the observation sequence of a specific area and the special payload configuration) is retained after abstraction. The abstracted data is reorganized to ensure that each record clearly reflects the remaining specific information. Database views or middleware can be used to achieve dynamic data reorganization. By abstracting the commonality part, the remaining data is more concise and clear, highlighting the uniqueness of each task; facilitating subsequent classification and labeling, improving data management and retrieval efficiency; while maintaining the separation between general data and specific operations, facilitating flexible combination and customized application.

[0129] The abstracted operation and control plan templates of the various types of aircrafts in different task scenarios are classified, and the classification information is labeled in the abstracted operation and control plan templates of the various types of aircrafts in different task scenarios.

[0130] Each operation and control plan template contains key parameter settings such as flight altitude, orbit parameters, and sensor working mode of the aircraft. Machine learning classification algorithms (such as decision tree and random forest) or rule-based classifiers are used to classify the abstracted data. Classification dimensions can include task type (imaging, communication, and obstacle avoidance, etc.), aircraft model, environmental conditions (light and weather, etc.), task priority, and task size, etc. The classification results are labeled as metadata to each operation and control plan template, forming structured label information. The labeled information can be used for quick retrieval, filtering, and aggregation analysis. Through classification and labeling, the fine management of operation and control plan templates is realized; efficient data query and filtering are supported, improving data utilization efficiency; classification information helps subsequent template matching and optimization, enhancing the response capability of the system to diversified task demands.

[0131] Based on the labeled operation and control plan templates of the various types of aircrafts in different task scenarios, a scheme template library is constructed.

[0132] The classified and labeled operation and control plan templates are integrated into the scheme template library, and each operation and control plan template corresponds to a typical task scenario or a specific aircraft model. The operation and control plan template content includes general data (commonality operation and parameter setting) and specific template (specific operation and parameter setting under specific task scenario). Distributed database or cloud storage system can be used to store the scheme template library, supporting version control and dynamic update. The functions of adding, deleting, modifying, and inquiring templates are realized, supporting user-defined template import and export. The effectiveness of the templates is evaluated regularly, and obsolete templates are updated or eliminated in time.

[0133] For S104, a multi-objective optimization algorithm is used to determine the operation control scheme based on the multiple alternative operation control plans, and the application provides an optional embodiment:

[0134] A multi-objective function is constructed based on the multiple alternative operation control plans.

[0135] A multi-objective function is constructed. Among them, f1(x) represents a task completion time minimization objective function, f2(x) represents an energy consumption minimization objective function, f3(x) represents a communication quality maximization objective function, and f4(x) represents a task execution safety maximization objective function.

[0136] A constraint condition is constructed.

[0137] A constraint condition is constructed. For the aircraft dynamics constraint g1(x), the following is obtained:

[0138] g1(x) = v max -|v(x)| ≥0;

[0139] For the energy constraint g2(x), the following is obtained:

[0140] g2(x) = E remaining - E(x)≥0;

[0141] For the communication coverage constraint g3(x), the following is obtained:

[0142] g3(x) = C(x) - C min ≥0;

[0143] For the task timing constraint g4(x), the following is obtained:

[0144] g4(x) = t end (x) - t deadline ≤0;

[0145] Where x represents a decision variable vector, including flight path parameters, communication frequency band parameters, attitude adjustment parameters, etc.; v max represents the maximum speed of the aircraft; v(x) represents the current speed of the aircraft; E remaining represents the remaining energy or remaining power of the aircraft; E(x) represents the power of the aircraft; C(x) represents the communication range of the aircraft; C min represents the minimum communication frequency band of the aircraft; t deadline represents the task deadline of the aircraft; t end (x) represents the task end time.

[0146] An iteration priority is obtained, and weights of the multi-objective function are adjusted based on the iteration priority and the basic information, to obtain an adjusted multi-objective function.

[0147] Based on the adaptive weight adjustment mechanism, the weights of the objective functions are adjusted according to the task priority and the basic information.

[0148] The improved non-dominated sorting genetic algorithm is used to solve the adjusted multi-objective function based on the constraint condition, to determine the operation and control scheme.

[0149] For solving the adjusted multi-objective function based on the constraint condition by using the improved non-dominated sorting genetic algorithm to determine the operation and control scheme, an optional embodiment of the present application is provided:

[0150] Let the value of k be 1.

[0151] An initial population is constructed based on the adjusted multi-objective function, and the initial population is taken as the population of the kth iteration; the initial population includes multiple individuals; the multiple individuals represent multiple solutions of the adjusted multi-objective function; and the multiple solutions represent multiple alternative operation and control plans.

[0152] The population of the kth iteration is quickly non-dominantly sorted, to obtain a non-dominantly sorted population of the kth iteration.

[0153] The crowding degree of each individual in the non-dominantly sorted population of the kth iteration is calculated, and the non-dominantly sorted population of the kth iteration is selected based on the constraint condition and the crowding degree of each individual, to obtain a selected population of the kth iteration.

[0154] The selected population of the kth iteration is crossbred and mutated by using tournament selection, simulated binary crossover, and polynomial mutation, to obtain a sub-population of the kth iteration.

[0155] The sub-population of the kth iteration is screened by using an elitist reservation strategy, to obtain a final population of the kth iteration.

[0156] If the iteration end condition is met, the individual in the final population of the kth iteration is output as the operation and control scheme.

[0157] The present application does not limit the iteration end condition, for example, the iteration end condition includes reaching a maximum iteration number or a convergence threshold.

[0158] If the iteration end condition is not met, the final population of the kth iteration is taken as the population of the (k+1)th iteration, the value of k is increased by 1, and the step of quickly non-dominantly sorting the population of the kth iteration to obtain the non-dominantly sorted population of the kth iteration is returned.

[0159] The above embodiment generates an initialization population based on multiple alternative operation control schemes, divides individuals in the population into layers according to a dominance relationship, and completes fast non-dominant sorting. The crowding distance of individuals in the same layer is calculated to maintain the diversity of the population. Selection, crossover and mutation operations are performed, and a new generation population is generated by using tournament selection, simulated binary crossover and polynomial mutation. The weights of each objective function are adjusted according to the task priority and basic information. An elite reservation strategy is used to ensure that excellent individuals in the parent generation directly enter the next generation, and the algorithm is terminated when the maximum iteration number or the convergence threshold is reached. Through the dynamic weight adjustment and constraint processing mechanism in the algorithm, the algorithm can better adapt to the complex and variable characteristics of low-orbit vehicles.

[0160] The above embodiment mainly uses the improved non-dominant sorting genetic algorithm to optimize the operation control scheme for multiple objective functions. In specific applications, the operation control scheme can also be adjusted according to the actual situation. In view of this, the present application provides an alternative embodiment:

[0161] According to the task demand information and the basic information of the vehicle, the parameters are calculated and optimized. First, the task demand information, basic information and operation control scheme are comprehensively analyzed to determine the type and priority of the parameters to be adjusted. The association rule mining algorithm is used to determine the small models to be called, and then the business models such as maneuver path planning, communication frequency band coverage calculation and vehicle attitude control are called to adjust the parameters in the operation control scheme. The large model interacts with the small model through a standardized interface protocol, and after the small model completes the calculation, the large model integrates the results using a multi-objective optimization algorithm and performs collaborative optimization. When the state quantity of the vehicle does not meet the task demand, according to the correlation analysis of the flight time and task execution priority of the vehicle by the large model, the maneuver path or task execution order is adjusted based on the calculation results of the business model to ensure that the task can be completed within the range allowed by the low-orbit vehicle capability. For example, if the current position of the vehicle deviates from the ideal position preset by the template, the flight angle and speed parameters that need to be adjusted are calculated according to the large model, and the flight path in the operation control scheme is corrected. At the same time, considering the time limit of the task and the energy consumption of the vehicle, the task execution priority and resource allocation relationship are analyzed by the large model, and the task execution order and device working mode parameters in the basic information are optimized.

[0162] The natural language processing capability of the large model is used to realize automatic and accurate analysis, and the structured information is quickly and accurately converted, which greatly improves the task analysis efficiency and accuracy; the multi-element characteristics are efficiently matched, which significantly improves the accuracy and efficiency of matching. The traditional small model such as the aircraft path planning calculation model and the communication frequency band calculation model is called to build a collaborative working mode of the large model and the traditional small model, realize the global optimization of the parameters, and meet the efficiency and accuracy requirements of the aircraft operation control task operation control scheme generation. In terms of human cost and development work, the application reduces the workload of manually generating and adjusting the operation control plan, and reduces the dependence on professional personnel. At the same time, the construction of the scheme template library and the application of the large model make the system have good universality and expansibility. When facing new task requirements or aircraft types, only the scheme template library needs to be updated and the large model needs to be fine-tuned, without the need for secondary development of related software systems, thereby reducing the development cost and difficulty. Compared with the prior art, the application realizes intelligent improvement and innovation, effectively solves the problems of low efficiency, insufficient accuracy, poor universality and expansibility of the operation control scheme generation, and can quickly and accurately generate an operation control plan that meets the diversified task requirements of low-orbit aircraft, thereby improving the task execution success rate and the overall operation efficiency of the aircraft.

[0163] Figure 2 A structural diagram of a control device of an aircraft provided by an embodiment of the application is shown in Figure 2 Based on the control method of the aircraft provided by the foregoing embodiment, the application also correspondingly provides a control device of an aircraft, which includes:

[0164] An acquisition module is configured to acquire task demand information, operation parameter information and basic information of the aircraft in real time; the basic information includes orbit position information, aircraft type information and equipment state information.

[0165] A comparison module is configured to compare the basic information with the scheme template library to obtain a target operation control plan template.

[0166] A filling module is configured to calculate and fill elements in the target operation control plan template based on the task demand information and the operation parameter information to obtain a plurality of alternative operation control plans.

[0167] An operation control scheme determination module is configured to determine an operation control scheme based on the plurality of alternative operation control plans by using a multi-objective optimization algorithm.

[0168] A control module is configured to control the aircraft based on the operation control scheme.

[0169] As an optional embodiment, the comparison module specifically includes:

[0170] A vector conversion unit is configured to perform vector conversion on the basic information to obtain a feature vector; the feature vector comprises a feature vector corresponding to the basic information.

[0171] A preprocessing unit is configured to perform normalization and standardization processing on the feature vector to obtain a processed feature vector.

[0172] A key vector unit is configured to dynamically calculate an association weight between the processed feature vectors based on an attention mechanism to obtain a key vector.

[0173] A comparison unit is configured to compare the key vector with a scheme template library based on a similarity matching algorithm to obtain a target operation and control plan template; the scheme template library comprises operation and control plan templates of various aircrafts in different task scenarios.

[0174] As an optional embodiment, the filling module specifically comprises:

[0175] A segmentation unit is configured to perform text segmentation on the task demand information to obtain a plurality of sub-information; the sub-information comprises a single word or phrase.

[0176] A mapping unit is configured to map each sub-information to obtain a vector corresponding to each sub-information.

[0177] A semantic association unit is configured to analyze the vectors corresponding to each sub-information to determine semantic associations between the sub-information, and obtain a key element corresponding to the task demand information.

[0178] A to-be-supplemented element set determination unit is configured to input the key element and the operation parameter information into an element model to obtain a plurality of to-be-supplemented element sets.

[0179] A supplement unit is configured to obtain a plurality of alternative operation and control plans based on the plurality of to-be-supplemented element sets and the target operation and control plan template; the number of to-be-supplemented element sets is consistent with the number of alternative operation and control plans.

[0180] As an optional embodiment, the device further comprises:

[0181] A historical data determination module is configured to obtain operation and control plan templates of various aircrafts in different task scenarios.

[0182] A data analysis module is configured to analyze the operation and control plan templates of various aircrafts in different task scenarios to determine common operations and parameter settings in the operation and control plan templates of various aircrafts in different task scenarios.

[0183] An abstract processing module is configured to abstract the common operations and the parameter settings into general data.

[0184] an abstracting module configured to abstract common operations and parameters corresponding to the general data from the operation and control plan templates of the various types of aircraft in different task scenarios, to obtain abstracted operation and control plan templates of the various types of aircraft in different task scenarios.

[0185] a classification labeling module configured to classify the abstracted operation and control plan templates of the various types of aircraft in different task scenarios, and label classification information in the abstracted operation and control plan templates of the various types of aircraft in different task scenarios.

[0186] a construction module configured to construct a scheme template library based on the labeled operation and control plan templates of the various types of aircraft in different task scenarios.

[0187] As an optional embodiment, the operation and control scheme determination module specifically includes:

[0188] a multi-objective function determination unit configured to construct a multi-objective function based on the multiple alternative operation and control plans.

[0189] a constraint condition determination unit configured to construct a constraint condition.

[0190] an adjustment unit configured to obtain a task priority, and adjust weights of the multi-objective function based on the task priority and the basic information, to obtain an adjusted multi-objective function.

[0191] a solution unit configured to solve the adjusted multi-objective function based on the constraint condition by using an improved non-dominated sorting genetic algorithm, to determine the operation and control scheme.

[0192] As an optional embodiment, the solution unit specifically includes:

[0193] a setting subunit configured to set a value of k to 1.

[0194] an initialization subunit configured to construct an initialization population based on the adjusted multi-objective function, and use the initialization population as a population of the kth iteration; the initialization population includes multiple individuals; the multiple individuals represent multiple solutions of the adjusted multi-objective function; the multiple solutions represent multiple alternative operation and control plans.

[0195] a sorting subunit configured to perform fast non-dominated sorting on the population of the kth iteration, to obtain a non-dominated sorted population of the kth iteration.

[0196] a selection subunit configured to calculate a crowding degree of each individual in the non-dominated sorted population of the kth iteration, and select the non-dominated sorted population of the kth iteration based on the constraint condition and the crowding degree of each individual, to obtain a selected population of the kth iteration.

[0197] crossing and mutating the selected population of the kth iteration to obtain a sub-population of the kth iteration.

[0198] screening the sub-population of the kth iteration to obtain a final population of the kth iteration.

[0199] The first judging sub-unit is configured to, if the iteration end condition is met, output the individual in the final population of the kth iteration as the operation and control scheme.

[0200] The second judging sub-unit is configured to, if the iteration end condition is not met, take the final population of the kth iteration as the population of the (k+1)th iteration, increase the value of k by 1, and return to the sorting sub-unit.

[0201] The embodiment of the present application provides a computer device, including a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor executes the computer program to realize the control method of the aircraft.

[0202] The embodiment of the present application provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the control method of the aircraft.

[0203] The embodiment of the present application provides a computer program product, including a computer program, and the computer program is executed by a processor to realize the control method of the aircraft.

[0204] It should be noted that each of the embodiments in the present specification is described in a progressive manner, and the same parts of each of the embodiments can be referred to each other, and each of the embodiments mainly describes the difference from other embodiments. Especially, the device and equipment embodiments are basically similar to the method embodiments, so the description is relatively simple, and the related parts can be referred to the part of the method embodiment. The device and equipment embodiments described above are only schematic, and the units described as separate components can be or can not be physically separated, and the components indicated as units can be or can not be physical components, that is, they can be located in one place, or can be distributed on multiple network components. According to the actual needs, part or all of the modules can be selected to realize the purpose of the embodiment scheme. Those skilled in the art can understand and implement without creative labor.

[0205] The above merely provides one specific embodiment of the present application, but the protection scope of the present application is not limited thereto, any changes or replacements within the technical scope disclosed by the present application, which can be easily thought by any person skilled in the art, should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A control method for an aircraft, characterized in that, The control method for the aircraft includes: The system acquires real-time mission requirements, operational parameters, and basic information of the spacecraft; the basic information includes orbital position information, spacecraft type information, and equipment status information. The basic information and the scheme template library are compared to obtain the target operation and control plan template; Based on the task requirement information and the operation parameter information, the elements in the target operation control plan template are calculated and filled to obtain multiple alternative operation control plans. A multi-objective optimization algorithm is used to determine the operation and control scheme based on the multiple alternative operation and control plans; The aircraft is controlled based on the aforementioned operational control scheme; The process involves calculating and filling in the elements of the target operation control plan template based on the task requirement information and the operational parameter information to obtain multiple alternative operation control plans, specifically including: The task requirement information is segmented into multiple sub-information items; each sub-information item includes a single word or phrase. Map each piece of information to obtain a vector corresponding to each piece of information; Analyze the vector corresponding to each sub-information to determine the semantic relationship between the sub-information, and obtain the key elements corresponding to the task requirement information; The key elements and the operating parameter information are input into the element model to obtain multiple sets of elements to be supplemented; Based on the multiple sets of elements to be supplemented and the target operation and control plan template, the multiple alternative operation and control plans are obtained; the number of the sets of elements to be supplemented and the number of alternative operation and control plans are the same.

2. The control method for an aircraft according to claim 1, characterized in that, The comparison of the basic information with the scheme template library to obtain the target operation and control plan template specifically includes: The basic information is transformed into a vector to obtain a feature vector; the feature vector includes the feature vector corresponding to the basic information. The feature vector is normalized and standardized to obtain the processed feature vector; The association weights between the processed feature vectors are dynamically calculated based on the attention mechanism to obtain the key vectors. The key vector is compared with the scheme template library based on the similarity matching algorithm to obtain the target operation and control plan template; the scheme template library includes operation and control plan templates for various types of aircraft under different mission scenarios.

3. The control method for an aircraft according to claim 1, characterized in that, The method further includes: Obtain operational control plan templates for various types of aircraft under different mission scenarios; The operation and control plan templates of various types of aircraft under different mission scenarios were analyzed to determine the common operations and parameter settings in the operation and control plan templates of various types of aircraft under different mission scenarios. The common operations and parameter settings are abstracted into general data; The common operations and parameter settings corresponding to the general data are extracted from the operation and control plan templates of various types of aircraft in different mission scenarios to obtain the extracted operation and control plan templates of various types of aircraft in different mission scenarios. The operation and control plan templates for various types of aircraft after extraction are classified under different mission scenarios, and the classification information is marked in the operation and control plan templates for various types of aircraft after extraction under different mission scenarios; A template library of operational control plans for various types of aircraft under different mission scenarios, based on the labeled templates, has been constructed.

4. The control method for an aircraft according to claim 1, characterized in that, The process of determining the operation and control scheme based on the multiple alternative operation and control plans using a multi-objective optimization algorithm specifically includes: A multi-objective function is constructed based on the aforementioned multiple alternative operation and control plans; Construct constraints; Obtain the task priority, and adjust the weights of the multi-objective function based on the task priority and the basic information to obtain the adjusted multi-objective function; An improved non-dominated sorting genetic algorithm is used to solve the adjusted multi-objective function based on the constraints to determine the operation and control scheme.

5. The control method for an aircraft according to claim 4, characterized in that, The improved non-dominated sorting genetic algorithm solves the adjusted multi-objective function based on the constraints to determine the operation and control scheme, specifically including: Let the value of k be 1; An initial population is constructed based on the adjusted multi-objective function, and this initial population is used as the population for the k-th iteration. The initial population includes multiple individuals, each representing a different solution to the adjusted multi-objective function. Each solution represents a different alternative operation and control plan. Perform fast non-dominated sorting on the population in the k-th iteration to obtain the population after non-dominated sorting in the k-th iteration; Calculate the crowding degree of each individual in the population after the non-dominated sorting of the k-th iteration, and select the population after the non-dominated sorting of the k-th iteration based on the constraints and the crowding degree of each individual to obtain the selected population after the k-th iteration. Tournament selection, simulated binary crossover, and polynomial mutation are used to perform crossover and mutation processing on the population selected in the k-th iteration to obtain the subpopulation in the k-th iteration. An elite retention strategy is used to screen the subpopulation of the k-th iteration to obtain the final population of the k-th iteration. If the iteration termination condition is met, then the individuals in the final population of the k-th iteration are output as the operation and control scheme. If the iteration termination condition is not met, the final population of the k-th iteration is used as the population of the (k+1)-th iteration, the value of k is incremented by 1, and the process of performing fast non-dominated sorting on the population of the k-th iteration is returned to obtain the population after non-dominated sorting in the k-th iteration.

6. A control device for an aircraft, characterized in that, The control device for the aircraft includes: The acquisition module is used to acquire the mission requirements, operating parameters, and basic information of the spacecraft in real time; the basic information includes orbital position information, spacecraft type information, and equipment status information. The comparison module is used to compare the basic information with the scheme template library to obtain the target operation and control plan template; The filling module is used to calculate and fill the elements in the target operation and control plan template based on the task requirement information and the operation parameter information to obtain multiple alternative operation and control plans; The operation control scheme determination module is used to determine the operation control scheme based on the multiple alternative operation control schemes using a multi-objective optimization algorithm. A control module is used to control the aircraft based on the aforementioned motion control scheme; The fill module specifically includes: The segmentation unit is used to perform text segmentation on the task requirement information to obtain multiple sub-information; the sub-information includes a single word or phrase; The mapping unit is used to map each piece of information to obtain a vector corresponding to each piece of information; The semantic association unit is used to analyze the vector corresponding to each sub-information to determine the semantic association between the sub-information and obtain the key elements corresponding to the task requirement information; The element set to be supplemented unit is used to input the key elements and the operating parameter information into the element model to obtain multiple element sets to be supplemented; The supplementary unit is used to obtain the multiple alternative operation and control plans based on the multiple sets of elements to be supplemented and the target operation and control plan template; the number of the sets of elements to be supplemented and the number of alternative operation and control plans are the same.

7. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the control method of the aircraft according to any one of claims 1-5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the control method of the aircraft as described in any one of claims 1-5.

9. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the control method of the aircraft as described in any one of claims 1-5.

Images