Unmanned helicopter terrain simulation flight method and system based on improved model predictive control algorithm

Abstract

The invention provides an unmanned helicopter terrain-imitated flight method and system based on an improved model predictive control algorithm. Relates to the technical field of unmanned flight control, and comprises a state modeling module, the state modeling module carries out synchronous modeling processing on flight initial state data of an unmanned helicopter, the state modeling module forms a digital twin model through the synchronous modeling processing, and the synchronous modeling processing adopts a state matrix generation method. According to the improved model predictive control algorithm-based land-imitated flight method and system for the unmanned helicopter, through multi-source data fusion and a dynamic weight updating mechanism, the influence of sensor errors and external disturbance on the flight control system is reduced, and the robustness and reliability of the system are enhanced. The optimized control algorithm improves the energy distribution efficiency, reduces the maintenance cost, and provides stronger guarantee for long-term stable operation of the unmanned helicopter.

Description

Technical Field

[0001] This invention relates to the field of unmanned flight control technology, specifically to a method and system for unmanned helicopters to follow the terrain for flight based on an improved model predictive control algorithm. Background Technology

[0002] Current low-altitude terrain-following flight of unmanned helicopters relies on onboard sensors for real-time perception of terrain undulations and altitude changes, using flight control algorithms to dynamically adjust attitude and throttle control. Control methods include Proportional-Integral-Derivative (PID) control, fuzzy control, and traditional Model Predictive Control (MPC). Flight trajectory constraints are achieved through the establishment of a dynamic model and real-time error feedback. Existing unmanned helicopters mostly employ a hierarchical control architecture, with attitude, velocity, and altitude loops operating independently, requiring manual tuning of control parameters. When facing complex terrain, sudden airflow, or ground disturbances, flight stability depends on sensor accuracy and control law robustness. This paper introduces feedforward control and Kalman filtering to mitigate the impact of measurement noise, and utilizes multi-source sensor fusion to improve attitude calculation accuracy.

[0003] However, existing technologies still have significant shortcomings. Current aircraft power management systems are inefficient; manual inspections and periodic maintenance consume considerable time and manpower, and the scope of inspection is limited, failing to cover the status of all equipment. The system often experiences response delays during operation, and fault signals cannot be identified and processed in a timely manner, easily leading to downtime or system damage. Data from each subsystem is independent, lacking a unified management and analysis platform, making global optimization of operational status impossible. Maintenance work relies on specialized technicians and testing equipment, resulting in high investment and long cycles. Different aircraft models have significantly different power requirements; existing system architectures are fixed, have poor adaptability, and cannot accommodate the operating environments of multiple aircraft types. This leads to low power distribution efficiency, reduced energy utilization, and limits the reliability and scalability of the flight control system. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a method and system for unmanned helicopters to follow terrain for flight based on an improved model predictive control algorithm. The technical problem this invention aims to solve is: how to form a digital twin model through synchronous modeling, and combine it with an improved model predictive control algorithm to predict and control the flight state, thereby achieving accurate and stable unmanned helicopters to follow terrain for flight.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an unmanned helicopter terrain-following flight system based on an improved model predictive control algorithm, comprising: a state modeling module, wherein the state modeling module performs synchronous modeling processing on the initial flight state data of the unmanned helicopter, and the state modeling module forms a digital twin model through the synchronous modeling processing, wherein the synchronous modeling processing adopts the state matrix generation method.

[0006] The data prediction module collects and predicts data on the flight status perception data of the unmanned helicopter. The data prediction module generates local predictive control commands through the collection and prediction processing, and the collection and prediction processing adopts an improved model predictive control algorithm.

[0007] The data fusion module uploads and fuses the local predictive control command and the state data of the digital twin model. The data fusion module forms a global state estimation result through the upload and fusion. The upload and fusion adopts time delay compensation and consistency correction. The global state estimation result includes attitude angle state parameters and environmental disturbance compensation parameters.

[0008] An attitude control module performs multi-objective coordinated control on the global state estimation results. The attitude control module generates terrain-following flight control commands through the multi-objective coordinated control. The multi-objective coordinated control employs an optimization solution method. The attitude control module drives the unmanned helicopter to adjust its actions based on the terrain-following flight control commands. The attitude control module generates execution response data through the action adjustments.

[0009] The feedback correction module performs feedback analysis on the execution response data and generates algorithm update data through the feedback analysis. The feedback analysis employs a dynamic weight update method.

[0010] Preferably, the initial flight state data includes sensor initial parameters and dynamic characteristic data. The state matrix generation method is based on dynamic equations and terrain mapping relationships. The dynamic equations adopt a nonlinear state coupling method. The terrain mapping relationship includes a terrain elevation matrix and a surface gradient matrix. The terrain elevation matrix and the surface gradient matrix are matched using spatial coordinate indices.

[0011] Preferably, the flight state perception data includes attitude angular velocity, barometric altitude, terrain distance, and environmental disturbance signals, and the improved model predictive control algorithm includes an adaptive weight adjustment mechanism and a constraint optimization mechanism, wherein the constraint optimization mechanism includes input constraints and state constraints.

[0012] Preferably, the adaptive weight adjustment mechanism employs a weighted predictive control method, and the model formula for the weighted predictive control method is as follows:

[0013] in, This is the local predictive control command vector, in units of N. This is the attitude control output, in N (unit: kilometres). For height control output, the unit is N. The target attitude angle is expressed in rad. The current attitude angle is expressed in rad. The desired height is in meters (m). This is the current altitude, in meters. Attitude proportional gain, in units of , This represents the attitude integral gain, in units of... , This is a high proportional gain, measured in N / m. High integral gain, in units of , The time integral of the attitude angle error, in units of , The time integral of the altitude error, in units of .

[0014] Preferably, the attitude angle state parameters include attitude angle, angular velocity and attitude stability index, and the environmental disturbance compensation parameters include airflow disturbance intensity coefficient and terrain gradient change coefficient.

[0015] Preferably, the delay compensation includes data sampling synchronization compensation and signal transmission delay compensation, and the consistency correction includes multi-source state matching correction and model parameter self-consistency correction.

[0016] Preferably, the optimization solution method includes attitude stabilization solution processing and height coordination solution processing, and the motion adjustment includes attitude adjustment and height tracking control motion adjustment.

[0017] Preferably, the attitude stabilization solution process generates an attitude control signal based on the attitude angle state parameters, the altitude coordination solution process generates an altitude correction signal based on the attitude control signal and the environmental disturbance compensation parameters, the optimization solution method fuses the attitude control signal and the altitude correction signal, and the attitude control module forms the terrain-following flight control command through the fusion.

[0018] Preferably, the dynamic weight update method includes adaptive energy consumption adjustment and deviation coupling correction. The adaptive energy consumption adjustment includes power change response adjustment and energy distribution balance adjustment, and the deviation coupling correction includes altitude deviation trend correction and attitude response gain correction.

[0019] The terrain-following flight method for unmanned helicopters based on improved model predictive control algorithms includes: S1. The initial flight state data of the unmanned helicopter is synchronously modeled to form a digital twin model. The synchronous modeling process adopts the state matrix generation method.

[0020] S2. The flight status perception data of the unmanned helicopter is collected and predicted to form local predictive control commands. The collection and prediction processing includes an adaptive weight adjustment mechanism and a constraint optimization mechanism.

[0021] S3. Upload and fuse the state data of the local predictive control command and the digital twin model to form a global state estimation result. The upload and fusion process includes delay compensation and consistency correction.

[0022] S4. Perform multi-objective coordinated control processing on the global state estimation results to form execution response data. The multi-objective coordinated control processing includes optimization solution methods and action adjustment mechanisms.

[0023] S5. The execution response data is subjected to feedback analysis and processing. The feedback correction module generates algorithm update data through the feedback analysis and processing. The feedback analysis and processing adopts the dynamic weight update method.

[0024] This invention provides a method and system for terrain-following flight of unmanned helicopters based on an improved model predictive control algorithm. It has the following beneficial effects: This invention, based on an improved model predictive control algorithm, enhances the flight stability of unmanned helicopters in complex terrain and unstable airflow environments. Employing a precise digital twin model and adaptive predictive control, the system can respond to terrain changes and external disturbances in real time, ensuring more accurate attitude and altitude control during flight and improving the aircraft's autonomy and adaptability.

[0025] This method and system for unmanned helicopter terrain-following flight, based on an improved model predictive control algorithm, reduces the impact of sensor errors and external disturbances on the flight control system through multi-source data fusion and a dynamic weight update mechanism, thereby enhancing the system's robustness and reliability. The optimized control algorithm improves energy distribution efficiency, reduces maintenance costs, and provides stronger assurance for the long-term stable operation of the unmanned helicopter. Attached Figure Description

[0026] Figure 1 A schematic diagram of the structure of an unmanned helicopter's terrain-following flight system; Figure 2 A flowchart illustrating the state modeling module; Figure 3 This is a flowchart illustrating the data prediction module process. Figure 4 This is a flowchart illustrating the data fusion module process. Figure 5 This is a flowchart of the attitude control module. Detailed Implementation

[0027] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] Example 1 like Figure 1-5 As shown, this embodiment of the invention provides a terrain-following flight system for an unmanned helicopter based on an improved model predictive control algorithm. The system includes a state modeling module, which synchronously models the initial flight state data of the unmanned helicopter. The state modeling module forms a digital twin model through this synchronous modeling process, which employs a state matrix generation method. The initial flight state data includes initial sensor parameters and dynamic characteristic data. The state matrix generation method is based on dynamic equations and terrain mapping relationships. The dynamic equations employ a nonlinear state coupling method. The terrain mapping relationship includes a terrain elevation matrix and a surface gradient matrix, which are matched using spatial coordinate indices.

[0029] The data prediction module collects and processes flight state perception data from the unmanned helicopter, generating local predictive control commands. This data collection and processing employs an improved model predictive control algorithm. Flight state perception data includes attitude angular velocity, pressure altitude, terrain distance, and environmental disturbance signals. The improved model predictive control algorithm includes an adaptive weight adjustment mechanism and a constraint optimization mechanism. The constraint optimization mechanism includes input constraints and state constraints. The adaptive weight adjustment mechanism uses a weighted predictive control method, and the model formula for the weighted predictive control method is as follows:

[0030] in, This is the local predictive control command vector, in units of N. This is the attitude control output, in N (unit: kilometres). For height control output, the unit is N. The target attitude angle is expressed in rad. The current attitude angle is expressed in rad. The desired height is in meters (m). This is the current altitude, in meters. Attitude proportional gain, in units of , This represents the attitude integral gain, in units of... , This is a high proportional gain, measured in N / m. High integral gain, in units of , The time integral of the attitude angle error, in units of , The time integral of the altitude error, in units of .

[0031] The data fusion module uploads and fuses the state data from local predictive control commands and the digital twin model. This fusion process generates a global state estimation result, employing time delay compensation and consistency correction. The global state estimation result includes attitude angle state parameters and environmental disturbance compensation parameters. Attitude angle state parameters include attitude angle, angular velocity, and attitude stability indices. Environmental disturbance compensation parameters include airflow disturbance intensity coefficient and terrain gradient change coefficient. Time delay compensation includes data sampling synchronization compensation and signal transmission delay compensation. Consistency correction includes multi-source state matching correction and model parameter self-consistency correction.

[0032] The attitude control module performs multi-objective coordinated control based on the global state estimation results. This multi-objective coordinated control generates terrain-following flight control commands, employing an optimization solution method. Based on these commands, the attitude control module drives the unmanned helicopter to adjust its actions, generating execution response data. The optimization solution method includes attitude stabilization processing and altitude coordination processing. Action adjustments include attitude adjustment and altitude tracking control action adjustment. Attitude stabilization processing generates attitude control signals based on attitude angle state parameters, while altitude coordination processing generates altitude correction signals based on the attitude control signals and environmental disturbance compensation parameters. The optimization solution method fuses the attitude control signals and altitude correction signals, and the attitude control module uses this fusion to generate terrain-following flight control commands.

[0033] Input data and global state estimation: The attitude control module receives the global state estimation results from the data fusion module. The results include the aircraft's attitude angles, angular velocity, altitude and their changes, as well as compensation parameters for external environmental disturbances. The current state is: attitude angle Φ = 0.05 rad, angular velocity p = 0.02 rad / s, altitude... =150 m, environmental disturbance compensation parameters: wind speed disturbance is Δv=1 m / s.

[0034] Attitude stabilization solution processing: Attitude control signals are generated from attitude angle state parameters. The attitude stabilization solution uses the following formula to generate the control signals:

[0035] Target attitude angle =0.1 rad, control gain =0.8 N / rad:

[0036] This represents the instruction from the attitude control module to adjust the roll angle.

[0037] Highly coordinated solution processing: The height coordination solution takes into account attitude control signals and external disturbance compensation parameters to generate a height correction signal. Target height =160 m, environmental disturbances have an effect of Δv=1 m / s on the aircraft. The altitude correction signal is generated by the following formula:

[0038] Among them, control gain =0.3 N / m and wind speed disturbance compensation gain =0.1 N / (m / s).

[0039] Substitute the data into the calculation:

[0040] This represents the altitude control module's correction of the aircraft's altitude.

[0041] Optimization of solution methods and signal fusion: In the optimization solution method, attitude control signal and height correction signal The signals will be fused to generate the final terrain-following flight control commands. The fusion method uses a weighted average or optimization algorithm to combine the two control signals. A simple weighted average method is used:

[0042] in, =0.6 and =0.4, indicating that attitude control has a higher weight than altitude control in the terrain-following flight control commands.

[0043] Substitute the data into the calculation:

[0044] The results indicate that the fused terrain-following flight control commands are used to drive the aircraft to adjust its attitude and altitude.

[0045] The feedback correction module performs feedback analysis on the execution response data, generating algorithm update data through this analysis. The feedback analysis employs a dynamic weight update method. This method includes adaptive energy consumption adjustment and deviation coupling correction. Adaptive energy consumption adjustment includes power change response adjustment and energy distribution balance adjustment, while deviation coupling correction includes altitude deviation trend correction and attitude response gain correction.

[0046] A method for unmanned helicopters to follow terrain for flight based on an improved model predictive control algorithm includes: S1. Synchronous modeling and processing of the initial flight state data of the unmanned helicopter to form a digital twin model. The synchronous modeling and processing adopts the state matrix generation method.

[0047] S2. Collect and predict data on the flight status of the unmanned helicopter to form local predictive control commands. The collection and prediction processing includes an adaptive weight adjustment mechanism and a constraint optimization mechanism.

[0048] S3. Upload and fuse the state data of the local predictive control command and the digital twin model to form a global state estimation result. The upload and fusion process includes time delay compensation and consistency correction.

[0049] S4. Perform multi-objective coordinated control processing on the global state estimation results to form execution response data. Multi-objective coordinated control processing includes optimization solution methods and action adjustment mechanisms.

[0050] S5. Feedback analysis and processing are performed on the execution response data. The feedback correction module generates algorithm update data through feedback analysis and processing, and the dynamic weight update method is used for feedback analysis and processing.

[0051] Example 2 This embodiment establishes dynamic equations using a nonlinear state coupling method, enabling precise modeling and control of the complex dynamic behavior of unmanned helicopters during flight.

[0052] 1. Establish the dynamic equations Describing the state changes of an unmanned helicopter during flight requires defining its state variables, including position, velocity, attitude angles, and angular velocity. It is necessary to consider the interactions between different dynamic parameters during flight, dividing the variables into several subsystems and describing them using nonlinear state coupling equations.

[0053] Let the state vector of the helicopter be... Where: x, y, z are position coordinates. ,θ, denoted as roll, pitch, and yaw angles, u, v, w as linear airspeeds, and p, q, r as roll, pitch, and yaw angular velocities.

[0054] The dynamic equations are established based on the motion equations of the unmanned helicopter and the external disturbances. The dynamic model of the helicopter is as follows:

[0055] in, It is a nonlinear function representing the coupling relationship between various state variables, where u is the input control signal and t is the time variable.

[0056] 2. Nonlinear Coupling Description The flight control system of a helicopter involves multiple nonlinear dynamic characteristics, such as aerodynamic properties, engine power output, and aircraft weight, resulting in complex nonlinear terms in the dynamic equations. The relationship between the aircraft's roll and pitch angles is not nonlinear but is strongly influenced by the airframe structure and aerodynamic characteristics. The nonlinear dynamic equations of a helicopter can be expressed as:

[0057]

[0058]

[0059]

[0060]

[0061]

[0062] in, , , These represent the helicopter's moment of inertia. The nonlinear dynamic equation model couples the various dynamic states through nonlinear terms, realistically reflecting the flight behavior of the unmanned helicopter.

[0063] 3. Terrain Influence Modeling By introducing terrain mapping relationships, the impact of terrain on flight is considered. Terrain information, including terrain elevation and surface gradient, is matched using spatial coordinate indexing and coupled with state variables in the dynamic model. The elevation data matrix and surface gradient matrix are used to describe the influence of terrain on UAV flight.

[0064] The terrain elevation matrix H(x,y) and the surface gradient matrix G(x,y) describe the terrain undulations of a specific flight area. Modeling the terrain parameters allows for dynamic adjustment of the helicopter's flight path and speed to avoid obstacles during low-altitude flight.

[0065] 4. Numerical Solution and Simulation After establishing the dynamic model, numerical simulation was used to solve the nonlinear equations. The Euler method was employed for numerical integration to obtain the dynamic response of the helicopter under different control inputs and flight environments. Simulation results under different terrain conditions were compared to verify the accuracy and reliability of the nonlinear coupled dynamic model.

[0066] In a certain simulation scenario, the initial state of the unmanned helicopter is as follows: .

[0067] The flight altitude is 100 meters, the helicopter is flying horizontally, and there is no initial yaw or pitch angle. Based on the known dynamic equations, numerical integration methods are used to simulate the changes in position, velocity, and attitude of the helicopter during flight.

[0068] The specific flight trajectory data is obtained through numerical integration:

[0069] This indicates that the helicopter's state at time t is: The flight position is (10,0,98)(10,0,98)(10,0,98).

[0070] The attitude angle and angular velocity are Φ=0.1°, θ=0.05°, and p=0, respectively.

[0071] Example 3 This embodiment generates local control commands through an adaptive weight adjustment mechanism and a constraint optimization mechanism, ensuring the stability and safety of the aircraft in complex flight environments.

[0072] 1. Adaptive weight adjustment mechanism The model formula for the weighted predictive control method is:

[0073] in, This is the local predictive control command vector, in units of N. This is the attitude control output, in N (unit: kilometres). For height control output, the unit is N. The target attitude angle is expressed in rad. The current attitude angle is expressed in rad. The desired height is in meters (m). This is the current altitude, in meters. Attitude proportional gain, in units of , This represents the attitude integral gain, in units of... , This is a high proportional gain, measured in N / m. High integral gain, in units of , The time integral of the attitude angle error, in units of , The time integral of the altitude error, in units of .

[0074] 2. Constraint Optimization Mechanism The constraint optimization mechanism includes input constraints and state constraints to ensure that flight control commands are within a reasonable range and to avoid over-control or unsafe flight states.

[0075] Input constraints: The output of control commands must be within the aircraft's capabilities, ensuring that each control variable does not exceed its maximum permissible value. State constraints: The aircraft's attitude angles, speed, altitude, and other state variables should be limited within safe flight ranges to prevent loss of control or dangerous conditions.

[0076] 3. Data Example At a certain moment, the aircraft's status data is as follows: Current attitude angle =0.05rad, target attitude angle =0.02rad, current altitude =150m, target height =160m.

[0077] Calculate control commands:

[0078] Substitute the specific controller parameters It is 0.8 , It is 0.5 , It is 0.3 N / m. It is 0.1 ,get:

[0079] When t=1:

[0080] The corresponding control commands are obtained through numerical integration, and the flight status of the aircraft is adjusted in real time.

[0081] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A terrain-following flight system for unmanned helicopters based on an improved model predictive control algorithm, characterized in that: include: The state modeling module performs synchronous modeling processing on the initial flight state data of the unmanned helicopter. The state modeling module forms a digital twin model through the synchronous modeling processing, which adopts the state matrix generation method. The data prediction module collects and predicts the flight status perception data of the unmanned helicopter. The data prediction module generates local predictive control commands through the collection and prediction processing. The collection and prediction processing adopts an improved model predictive control algorithm. The data fusion module uploads and fuses the local predictive control command and the state data of the digital twin model. The data fusion module forms a global state estimation result through the upload and fusion. The upload and fusion adopts time delay compensation and consistency correction. The global state estimation result includes attitude angle state parameters and environmental disturbance compensation parameters. An attitude control module performs multi-objective coordinated control on the global state estimation results, generates terrain-following flight control commands through the multi-objective coordinated control, employs an optimization solution method, drives the unmanned helicopter to adjust its actions based on the terrain-following flight control commands, and generates execution response data through the action adjustments. The feedback correction module performs feedback analysis on the execution response data and generates algorithm update data through the feedback analysis. The feedback analysis employs a dynamic weight update method.

2. The unmanned helicopter terrain-following flight system based on the improved model predictive control algorithm according to claim 1, characterized in that: The initial flight state data includes sensor initial parameters and dynamic characteristic data. The state matrix generation method is based on dynamic equations and terrain mapping relationships. The dynamic equations adopt a nonlinear state coupling method. The terrain mapping relationship includes a terrain elevation matrix and a surface gradient matrix. The terrain elevation matrix and the surface gradient matrix are matched using spatial coordinate indices.

3. The unmanned helicopter terrain-following flight system based on the improved model predictive control algorithm according to claim 1, characterized in that: The flight state perception data includes attitude angular velocity, barometric altitude, terrain distance, and environmental disturbance signals. The improved model predictive control algorithm includes an adaptive weight adjustment mechanism and a constraint optimization mechanism. The constraint optimization mechanism includes input constraints and state constraints.

4. The unmanned helicopter terrain-following flight system based on the improved model predictive control algorithm according to claim 3, characterized in that: The adaptive weight adjustment mechanism employs a weighted predictive control method, the model formula of which is: ， in, For local predictive control command vectors, For attitude control output. For high-level control output, For the target attitude angle, The current attitude angle, For the desired height, At the current altitude, For attitude scaling gain, For attitude integral gain, For high proportional gain, For high integral gain, The time integral of the attitude angle error. The time integral of the height error.

5. The unmanned helicopter terrain-following flight system based on the improved model predictive control algorithm according to claim 1, characterized in that: The attitude angle state parameters include attitude angle, angular velocity and attitude stability index, and the environmental disturbance compensation parameters include airflow disturbance intensity coefficient and terrain gradient change coefficient.

6. The unmanned helicopter terrain-following flight system based on the improved model predictive control algorithm according to claim 1, characterized in that: The delay compensation includes data sampling synchronization compensation and signal transmission delay compensation, and the consistency correction includes multi-source state matching correction and model parameter self-consistency correction.

7. The unmanned helicopter terrain-following flight system based on the improved model predictive control algorithm according to claim 1, characterized in that: The optimization solution method includes attitude stabilization solution processing and height coordination solution processing, and the motion adjustment includes attitude adjustment and height tracking control motion adjustment.

8. The unmanned helicopter terrain-following flight system based on the improved model predictive control algorithm according to claim 7, characterized in that: The attitude stabilization solution process generates an attitude control signal based on the attitude angle state parameters. The altitude coordination solution process generates an altitude correction signal based on the attitude control signal and the environmental disturbance compensation parameters. The optimization solution method fuses the attitude control signal and the altitude correction signal. The attitude control module forms the terrain-following flight control command through the fusion.

9. The unmanned helicopter terrain-following flight system based on the improved model predictive control algorithm according to claim 1, characterized in that: The dynamic weight update method includes adaptive energy consumption adjustment and deviation coupling correction. The adaptive energy consumption adjustment includes power change response adjustment and energy distribution balance adjustment. The deviation coupling correction includes altitude deviation trend correction and attitude response gain correction.

10. A method for unmanned helicopters to follow terrain for flight based on an improved model predictive control algorithm, characterized in that: include: S1. Synchronously model the initial flight state data of the unmanned helicopter to form a digital twin model. The synchronous modeling process adopts the state matrix generation method. S2. The flight state perception data of the unmanned helicopter is collected and predicted to form local predictive control commands. The collection and prediction processing includes an adaptive weight adjustment mechanism and a constraint optimization mechanism. S3. Upload and fuse the local predictive control command and the state data of the digital twin model to form a global state estimation result. The upload and fuse process includes time delay compensation and consistency correction. S4. Perform multi-objective coordinated control processing on the global state estimation results to form execution response data. The multi-objective coordinated control processing includes optimization solution methods and action adjustment mechanisms. S5. The execution response data is subjected to feedback analysis and processing. The feedback correction module generates algorithm update data through the feedback analysis and processing. The feedback analysis and processing adopts the dynamic weight update method.

Images