Rocket trajectory optimization and real-time adjustment system based on artificial intelligence

Abstract

The present invention relates to the field of artificial intelligence, and specifically refers to an artificial intelligence-based rocket trajectory optimization and real-time adjustment system, comprising a trajectory prediction module, a real-time adjustment control module, and an intelligent monitoring module. The present invention introduces a jump mechanism into the trajectory prediction module to effectively reduce the interference of intermediate states, directly focus on key time steps, and introduces forward diffusion and inverse process denoising mechanisms to effectively reduce the model's sensitivity to noise. In the real-time adjustment control module, based on model predictive control, the rocket's attitude and thrust are automatically adjusted, the state deviation is minimized by optimizing the objective function, and a constrained quadratic programming problem is adopted to make the solution process more efficient, reduce computational complexity, accelerate decision-making, significantly improve the rocket's trajectory optimization and real-time adjustment capabilities, and provide strong support for the realization of complex space missions.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence, and specifically to an artificial intelligence-based rocket trajectory optimization and real-time adjustment system. Background Art

[0002] With the development of artificial intelligence, its application in the field of rockets is gradually deepening, but general rocket trajectory optimization systems have problems such as low prediction accuracy and poor trajectory generation quality. There is a lack of denoising mechanism, and the model is more sensitive to noise and outliers in the input data, which may lead to inaccurate prediction results. At the same time, there is a lack of effective initialization and denoising means. The generated multimodal future trajectory may be of poor quality and cannot accurately reflect the real situation, affecting subsequent planning and execution; the adjustment part in ordinary rocket trajectory optimization systems has problems such as large state deviation and high computational complexity. The actual state of the rocket may deviate greatly from the expected state, resulting in inaccurate trajectory and complex solution process, which increases computing time and resource consumption and delays the decision-making process. Summary of the Invention

[0003] In response to the above situation, the present invention provides a rocket trajectory optimization and real-time adjustment system based on artificial intelligence. To address the problems of low prediction accuracy and poor trajectory generation quality, the present invention introduces a jump mechanism in the trajectory prediction module to effectively reduce the interference of intermediate states, directly focus on key time steps, and reduce the impact of noise, thereby improving the accuracy of trajectory prediction. The jump initializer uses the denoised deterministic trajectory as a basis to generate a more accurate multimodal future trajectory, improve the quality of trajectory planning, and introduces the forward diffusion and inverse process denoising mechanisms to effectively reduce the model's sensitivity to noise and improve the stability of the prediction results. To address the problems of large state deviation and high computational complexity in the adjustment part, the present invention automatically adjusts the rocket attitude and thrust based on model predictive control in the real-time adjustment control module. By optimizing the objective function and minimizing the state deviation, the use of thrust can also be effectively controlled. The constrained quadratic programming problem is adopted to make the solution process more efficient, reduce computational complexity, speed up decision-making, and significantly improve the rocket's trajectory optimization and real-time adjustment capabilities, providing strong support for the realization of complex space missions.

[0004] The artificial intelligence-based rocket trajectory optimization and real-time adjustment system provided by the present invention includes a trajectory prediction module, a real-time adjustment control module, and an intelligent monitoring module, specifically including the following:

[0005] The trajectory prediction module predicts the optimal flight trajectory of the rocket in real time by building a trajectory prediction model;

[0006] The real-time adjustment control module automatically adjusts the rocket's attitude and thrust based on model predictive control to ensure real-time optimization of the flight trajectory;

[0007] The intelligent monitoring module provides real-time visual display of the flight process, helping ground control personnel to monitor the flight trajectory and status at any time.

[0008] Furthermore, a trajectory prediction model is constructed in the trajectory prediction planning module, which specifically includes the following steps:

[0009] Step S1: Collect historical data, including the rocket's historical trajectory data and historical environmental data. The historical trajectory data includes position, velocity, acceleration, and attitude angle, and the historical environmental data includes wind speed, wind direction, air pressure, and temperature.

[0010] Step S2: Data preprocessing and feature extraction: preprocess the rocket's historical trajectory data and historical environmental data, including outlier processing, missing value filling, and data denoising. A dynamic attention mechanism is used to extract key moment features in the rocket's trajectory and generate a high-dimensional spatiotemporal feature vector.

[0011] Step S3: Introduce a pattern initializer to generate the starting state of the future trajectory through the high-dimensional spatiotemporal feature vector;

[0012] Step S4: The multimodal trajectory prediction unit generates a multimodal future trajectory by constructing a trajectory prediction model according to the starting state of the future trajectory;

[0013] Step S5: Introduce an adaptive reasoning accelerator and optimize the diffusion model in step S4 by using a recurrent neural network as the policy network to build an adaptive reasoning accelerator.

[0014] Step S6: collecting real-time data, collecting real-time flight data of the rocket;

[0015] Step S7: Dynamic adjustment: Introduce the reinforcement learning Adam optimizer to analyze the real-time flight data of the rocket and adjust the trajectory prediction model through the real-time feedback mechanism;

[0016] Furthermore, step S3 specifically includes the following contents:

[0017] Building a domain adaptation module: Based on rocket domain expertise, including weather changes, flight patterns, and thrust adjustments, the module is built. Preprocessed historical environmental data is fed into the module, which learns from the rocket's performance in different environments and adjusts the weights of the starting states for generating future trajectories.

[0018] Constructing a pattern initializer: Using contrastive learning and combining it with a domain adaptation module, we construct a pattern initializer to generate the starting state of future trajectories.

[0019] Furthermore, step S4 specifically includes the following steps:

[0020] Step S41: Input initial conditions, collect the rocket's initial position, speed, angle, mass, thrust and environmental factors, and generate the initial state of the trajectory;

[0021] Step S42: Forward diffusion, using Represents the initial state of the trajectory, introduces a diffusion model, and uses Gaussian noise to add noise to the initial state of the trajectory. The formula for adding noise is as follows:

[0022] ;

[0023] in, is the attenuation coefficient, is the time step, is the standard normally distributed noise;

[0024] Step S43: Generate trajectory distribution by gradually adding time steps , get the future trajectory distribution of multiple modes ;

[0025] Step S44: Inverse process, remove noise from the future trajectory distribution of multiple modes to obtain a deterministic future trajectory ;

[0026] Step S44 specifically includes the following contents:

[0027] For the future trajectory distribution of multiple modes, Start reverse denoising and use the inverse diffusion equation to update the trajectory. The inverse diffusion equation used is as follows:

[0028] ;

[0029] in, To adjust the parameters of denoising strength, is the denoising function;

[0030] Gradually reduce the time step value until , get the future trajectory distribution of the maximum number of modes after denoising;

[0031] Step S45: Introduce a jump mechanism to skip the intermediate moments and set the jump step size. In the forward diffusion and reverse process, skip The time step length is skipped as follows:

[0032] ;

[0033] in, is the time derivative of the trajectory state;

[0034] Step S46: Jump initializer, using the deterministic future trajectory generated in step S44 As a basis, initialize the multimodal future trajectory mode;

[0035] Step S47: A denoising unit denoises the initialized multimodal future trajectory modality, enhances the denoising effect through residual connections, and obtains a noise-free multimodal future trajectory.

[0036] Furthermore, step S44 specifically includes the following contents:

[0037] For the future trajectory distribution of multiple modes, Start reverse denoising and use the inverse diffusion equation to update the trajectory. The inverse diffusion equation used is as follows:

[0038] ;

[0039] in, To adjust the parameters of denoising strength, is the denoising function;

[0040] Gradually reduce the time step value until , and obtain the future trajectory distribution of the maximum number of modes after denoising.

[0041] Furthermore, in the real-time adjustment control module, the rocket attitude and thrust are automatically adjusted based on model predictive control, specifically including the following steps:

[0042] Step Q1: Build a model. Use the rocket's physical properties and Newton's laws of motion to build a state-space model. The state variables include the rocket's position, velocity, acceleration, and attitude angle. The input variables are thrust and attitude adjustment. The state equation is as follows:

[0043] ;

[0044] The output equation is as follows:

[0045] ;

[0046] in, is the state vector, is the vector containing all state variables, is the control input vector, which is a vector containing all input variables, is the output, and are process noise and measurement noise;

[0047] Step Q2: Objective function and constraint setting. The objective function is set to optimize the minimization of state deviation and control input usage. At the same time, state and input constraints are defined. The objective function formula is as follows:

[0048] ;

[0049] ;

[0050] ;

[0051] ;

[0052] ;

[0053] in, For time The state vector at time , is the desired state vector, Indicates The control input vector at time , 、 are the minimum and maximum allowed values of each state variable in the state vector, 、 To control the minimum and maximum allowed values of each input variable in the input vector;

[0054] Step Q3: Optimize the problem construction and determine the prediction time range , combining the objective function and constraints to solve the optimal control input, forming a constrained quadratic programming problem, the form used is as follows:

[0055] ;

[0056] ;

[0057] ;

[0058] Step Q4: Solve the optimization problem and use the initialization strategy to obtain the optimal thrust and attitude adjustment;

[0059] Step Q5: Apply the control input to adjust the thrust and attitude of the rocket according to the optimal thrust and attitude adjustment amounts.

[0060] The beneficial effects achieved by the present invention using the above scheme are as follows:

[0061] (1) To address the problems of low prediction accuracy and poor trajectory generation quality, the present invention introduces a jump mechanism into the trajectory prediction module to effectively reduce the interference of intermediate states, directly focus on key time steps, reduce the influence of noise, and thus improve the accuracy of trajectory prediction. The jump initializer uses the denoised deterministic trajectory as a basis to generate more accurate multimodal future trajectories, improve the quality of trajectory planning, and introduces the forward diffusion and inverse process denoising mechanisms to effectively reduce the model's sensitivity to noise and improve the stability of the prediction results.

[0062] (2) In order to solve the problems of large state deviation and high computational complexity in the adjustment part, the present invention automatically adjusts the rocket attitude and thrust based on model predictive control in the real-time adjustment control module. By optimizing the objective function and minimizing the state deviation, the use of thrust can be effectively controlled. The constrained quadratic programming problem is adopted to make the solution process more efficient, reduce the computational complexity, speed up the decision-making speed, and significantly improve the trajectory optimization and real-time adjustment capabilities of the rocket, providing strong support for the realization of complex space missions.

[0063] Figure 1 Schematic diagram of the artificial intelligence-based rocket trajectory optimization and real-time adjustment system provided by the present invention.

[0064] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention. DETAILED DESCRIPTION

[0065] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments; based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0066] In the description of the present invention, it should be understood that terms such as "upper", "lower", "front", "back", "left", "right", "top", "bottom", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the accompanying drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operated in a specific direction. Therefore, they should not be understood as limiting the present invention.

[0067] In the first embodiment, the present invention provides an artificial intelligence-based rocket trajectory optimization and real-time adjustment system, which includes a trajectory prediction module, a real-time adjustment control module, and an intelligent monitoring module, specifically including the following:

[0068] The trajectory prediction module predicts the optimal flight trajectory of the rocket in real time by building a trajectory prediction model;

[0069] The real-time adjustment control module automatically adjusts the rocket's attitude and thrust based on model predictive control to ensure real-time optimization of the flight trajectory;

[0070] The intelligent monitoring module provides real-time visual display of the flight process, helping ground control personnel to monitor the flight trajectory and status at any time.

[0071] Example 2: Based on the above example, this example constructs a trajectory prediction model in the trajectory prediction planning module, specifically including the following steps:

[0072] Step S1: Collect historical data, including the rocket's historical trajectory data and historical environmental data. The historical trajectory data includes position, velocity, acceleration, and attitude angle, and the historical environmental data includes wind speed, wind direction, air pressure, and temperature.

[0073] Step S2: Data preprocessing and feature extraction: preprocess the rocket's historical trajectory data and historical environmental data, including outlier processing, missing value filling, and data denoising. A dynamic attention mechanism is used to extract key moment features in the rocket's trajectory and generate a high-dimensional spatiotemporal feature vector.

[0074] Step S3: Introduce a pattern initializer to generate the starting state of the future trajectory through the high-dimensional spatiotemporal feature vector;

[0075] Step S4: The multimodal trajectory prediction unit generates a multimodal future trajectory by constructing a trajectory prediction model according to the starting state of the future trajectory;

[0076] Step S5: Introduce an adaptive reasoning accelerator and optimize the diffusion model in step S4 by using a recurrent neural network as the policy network to build an adaptive reasoning accelerator.

[0077] Step S6: collecting real-time data, collecting real-time flight data of the rocket;

[0078] Step S7: Dynamic adjustment: introduce the reinforcement learning Adam optimizer to analyze the real-time flight data of the rocket and adjust the trajectory prediction model through the real-time feedback mechanism.

[0079] In this embodiment, in step S2, statistical methods are used to identify and remove obvious outliers in the data set, interpolation is applied to fill missing trajectory data, a low-pass filter is used to denoise the flight data to ensure the smoothness of the data, and a dynamic attention mechanism is used to construct a dynamic attention model to focus on extracting key moment features and generate high-dimensional spatiotemporal feature vectors to capture important changes in the flight process.

[0080] In this embodiment, a recurrent neural network is used as the policy network to optimize the trajectory prediction model. By training the adaptive inference accelerator, the model can quickly adjust the prediction strategy under different environmental conditions to improve real-time performance.

[0081] Embodiment 3: This embodiment is based on the above embodiment. Step S3 specifically includes the following contents:

[0082] Building a domain adaptation module: Based on rocket domain expertise, including weather changes, flight patterns, and thrust adjustments, the module is built. Preprocessed historical environmental data is fed into the module, which learns from the rocket's performance in different environments and adjusts the weights of the starting states for generating future trajectories.

[0083] Constructing a pattern initializer: Using contrastive learning methods combined with a domain adaptation module, we construct a pattern initializer to generate the starting state of future trajectories.

[0084] Embodiment 4: This embodiment is based on the above embodiment, and step S4 specifically includes the following steps:

[0085] Step S41: Input initial conditions, collect the rocket's initial position, speed, angle, mass, thrust and environmental factors, and generate the initial state of the trajectory;

[0086] Step S42: Forward diffusion, using Represents the initial state of the trajectory, introduces a diffusion model, and uses Gaussian noise to add noise to the initial state of the trajectory. The formula for adding noise is as follows:

[0087] ;

[0088] in, is the attenuation coefficient, is the time step, is the noise of standard normal distribution;

[0089] Step S43: Generate trajectory distribution by gradually adding time steps , get the future trajectory distribution of multiple modes ;

[0090] Step S44: Inverse process, remove noise from the future trajectory distribution of multiple modes to obtain a deterministic future trajectory ;

[0091] Step S44 specifically includes the following contents:

[0092] For the future trajectory distribution of multiple modes, Start reverse denoising and use the inverse diffusion equation to update the trajectory. The inverse diffusion equation used is as follows:

[0093] ;

[0094] in, To adjust the parameters of denoising strength, is the denoising function;

[0095] Gradually reduce the time step value until , get the future trajectory distribution of the maximum number of modes after denoising;

[0096] Step S45: Introduce a jump mechanism to skip the intermediate moments and set the jump step size. In the forward diffusion and reverse process, skip The time step length is skipped as follows:

[0097] ;

[0098] in, is the time derivative of the trajectory state;

[0099] Step S46: Jump initializer, using the deterministic future trajectory generated in step S44 As a basis, initialize the multimodal future trajectory mode;

[0100] Step S47: A denoising unit denoises the initialized multimodal future trajectory modality, enhances the denoising effect through residual connections, and obtains a noise-free multimodal future trajectory.

[0101] In this embodiment, the initial conditions are set as:

[0102] Initial position;

[0103] Initial velocity;

[0104] The attitude angle is 0, 0, 0;

[0105] Mass is 15,000 kg;

[0106] The thrust is 300,000N;

[0107] Environmental factors: wind speed = 5 m / s, wind direction = 30 degrees, air pressure = 101325 Pa, temperature = 288 Kelvin;

[0108] The initial state of the generated trajectory is [1000, 2000, 3000, 50, 60, 70, 0, 0, 0, 15000, 300000, 5, 30, 101325, 288];

[0109] In the forward diffusion, the attenuation coefficient is set to 0.9, and the generated Gaussian noise is [0.5, -0.2, 0.1, 0, 0.3, -0.4, 0, 0, 0, 0, 0, 0, 0, 0, 0];

[0110] Add noise calculation, , perform element-by-element calculation and get is: [948.85,1899.92,2849.99,47.44,59.24,67.56,0,0,0,15000,300000,5,30,101325,288];

[0111] Assuming the time step 1 second, generating 3 modes ;

[0112] The generated trajectory mode is:

[0113] =[1005,2000,3001,50.1,60.5,69.8,0,0,0,15002,300001,5.1,30,101320,288];

[0114] =[1002,1998,3003,49.5,59.9,70.1,0,0,0,15001,300002,5.2,30,101325,288];

[0115] =[1001,2001,3000,50.2,60.1,69.5,0,0,0,15003,300000,5,30,101300,288];

[0116] Set to 0.1, denoising function , the trajectory after denoising is:

[0117] ;

[0118] Assuming jump step length 2 seconds, ;

[0119] =[1004,2001,3000,50,60,70,0,0,0,15000,300000,5,30,101325,288];

[0120] The final generated modal is:

[0121] =[1004,2001,3000,50,60,70,0,0,0,15000,300000,5,30,101325,288];

[0122] =[1003,2000,3001,50,60,70,0,0,0,15001,300001,5,30,101320,288];

[0123] =[1005,2002,3002,50,60,70,0,0,0,15002,300002,5,30,101330,288]

[0124] The above steps show how to generate the future trajectory of the rocket through specific parameter definitions and calculations.

[0125] Example 5, based on the above example, in the real-time adjustment control module, the rocket attitude and thrust are automatically adjusted based on model predictive control, specifically including the following steps:

[0126] Step Q1: Build a model. Use the rocket's physical properties and Newton's laws of motion to build a state-space model. The state variables include the rocket's position, velocity, acceleration, and attitude angle. The input variables are thrust and attitude adjustment. The state equation is as follows:

[0127] ;

[0128] The output equation is as follows:

[0129] ;

[0130] in, is the state vector, is the vector containing all state variables, is the control input vector, which is a vector containing all input variables, is the output, and are process noise and measurement noise;

[0131] Step Q2: Objective function and constraint setting. The objective function is set to optimize the minimization of state deviation and control input usage. At the same time, state and input constraints are defined. The objective function formula is as follows:

[0132] ;

[0133] ;

[0134] ;

[0135] ;

[0136] ;

[0137] in, For time The state vector at time , is the desired state vector, Indicates The control input vector at time , 、 are the minimum and maximum allowed values of each state variable in the state vector, 、 To control the minimum and maximum allowed values of each input variable in the input vector;

[0138] Step Q3: Optimize the problem construction and determine the prediction time range , combining the objective function and constraints to solve the optimal control input, forming a constrained quadratic programming problem, the form used is as follows:

[0139] ;

[0140] ;

[0141] ;

[0142] Step Q4: Solve the optimization problem and use the initialization strategy to obtain the optimal thrust and attitude adjustment;

[0143] Step Q5: Apply the control input to adjust the thrust and attitude of the rocket according to the optimal thrust and attitude adjustment amounts.

[0144] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0145] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

[0146] The present invention and its embodiments are described above. This description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if a person skilled in the art is inspired by this and, without departing from the purpose of the present invention, designs structures and embodiments similar to this technical solution without inventiveness, they shall fall within the scope of protection of the present invention.

Claims

1. Artificial intelligence-based rocket trajectory optimization and real-time adjustment system, characterized by: It includes trajectory prediction module, real-time adjustment control module and intelligent monitoring module, specifically including the following contents: The trajectory prediction module predicts the optimal flight trajectory of the rocket in real time by building a trajectory prediction model; The real-time adjustment control module automatically adjusts the rocket's attitude and thrust based on model predictive control to ensure real-time optimization of the flight trajectory; The intelligent monitoring module provides real-time visual display of the flight process, helping ground control personnel to monitor the flight trajectory and status at any time; Constructing a trajectory prediction model in the trajectory prediction planning module includes the following steps: Step S1: Collect historical data, including the rocket's historical trajectory data and historical environmental data. The historical trajectory data includes position, velocity, acceleration, and attitude angle, and the historical environmental data includes wind speed, wind direction, air pressure, and temperature. Step S2: Data preprocessing and feature extraction: preprocess the rocket's historical trajectory data and historical environmental data, including outlier processing, missing value filling, and data denoising. A dynamic attention mechanism is used to extract key moment features in the rocket's trajectory and generate a high-dimensional spatiotemporal feature vector. Step S3: Introduce a pattern initializer to generate the starting state of the future trajectory through the high-dimensional spatiotemporal feature vector; Step S4: The multimodal trajectory prediction unit generates a multimodal future trajectory by constructing a trajectory prediction model according to the starting state of the future trajectory; Step S5: Introduce an adaptive reasoning accelerator and optimize the diffusion model in step S4 by using a recurrent neural network as the policy network to build an adaptive reasoning accelerator. Step S6: collecting real-time data, collecting real-time flight data of the rocket; Step S7: Dynamic adjustment: introduce the reinforcement learning Adam optimizer to analyze the real-time flight data of the rocket and adjust the trajectory prediction model through the real-time feedback mechanism.

2. The artificial intelligence-based rocket trajectory optimization and real-time adjustment system according to claim 1, characterized in that: Step S3 specifically includes the following contents: Building a domain adaptation module: Based on rocket domain expertise, including weather changes, flight patterns, and thrust adjustments, the module is built. Preprocessed historical environmental data is fed into the module, which learns from the rocket's performance in different environments and adjusts the weights of the starting states for generating future trajectories. Constructing a pattern initializer: Using contrastive learning methods combined with a domain adaptation module, we construct a pattern initializer to generate the starting state of future trajectories.

3. The artificial intelligence-based rocket trajectory optimization and real-time adjustment system according to claim 1, characterized in that: Step S4 specifically includes the following steps: Step S41: Input initial conditions, collect the rocket's initial position, speed, angle, mass, thrust and environmental factors, and generate the initial state of the trajectory; Step S42: Forward diffusion, using Represents the initial state of the trajectory, introduces a diffusion model, and uses Gaussian noise to add noise to the initial state of the trajectory. The formula for adding noise is as follows: ； in, is the attenuation coefficient, is the time step, is the noise of standard normal distribution; Step S43: Generate trajectory distribution by gradually adding time steps , get the future trajectory distribution of multiple modes ; Step S44: Inverse process, remove noise from the future trajectory distribution of multiple modes to obtain a deterministic future trajectory ; Step S44 specifically includes the following contents: For the future trajectory distribution of multiple modes, Start reverse denoising and use the inverse diffusion equation to update the trajectory. The inverse diffusion equation used is as follows: ； in, To adjust the parameters of denoising strength, is the denoising function; Gradually reduce the time step value until , get the future trajectory distribution of the maximum number of modes after denoising; Step S45: Introduce a jump mechanism to skip the intermediate moments and set the jump step size. In the forward diffusion and reverse process, skip The time step length is skipped as follows: ； in, is the time derivative of the trajectory state; Step S46: Jump initializer, using the deterministic future trajectory generated in step S44 As a basis, initialize the multimodal future trajectory mode; Step S47: A denoising unit denoises the initialized multimodal future trajectory modality, enhances the denoising effect through residual connections, and obtains a noise-free multimodal future trajectory.

4. The artificial intelligence-based rocket trajectory optimization and real-time adjustment system according to claim 1, characterized in that: In the real-time adjustment control module, the rocket attitude and thrust are automatically adjusted based on model predictive control, specifically including the following steps: Step Q1: Build a model. Use the rocket's physical properties and Newton's laws of motion to build a state-space model. The state variables include the rocket's position, velocity, acceleration, and attitude angle. The input variables are thrust and attitude adjustment. The state equation is as follows: ； The output equation is as follows: ； in, is the state vector, is the vector containing all state variables, is the control input vector, which is a vector containing all input variables, is the output, and are process noise and measurement noise; Step Q2: Objective function and constraint setting. The objective function is set to optimize the minimization of state deviation and control input usage. At the same time, state and input constraints are defined. The objective function formula is as follows: ； ； ； ； ； in, For time The state vector at time , is the desired state vector, Indicates The control input vector at time , 、 are the minimum and maximum allowed values of each state variable in the state vector, 、 To control the minimum and maximum allowed values of each input variable in the input vector; Step Q3: Optimize the problem construction and determine the prediction time range , combining the objective function and constraints to solve the optimal control input, forming a constrained quadratic programming problem, the form used is as follows: ； ； ； Step Q4: Solve the optimization problem and use the initialization strategy to obtain the optimal thrust and attitude adjustment; Step Q5: Apply the control input to adjust the thrust and attitude of the rocket according to the optimal thrust and attitude adjustment amounts.

Images