A fault-tolerant adaptive deorbiting closed-loop control method and system for rocket final stage faults
By constructing an adaptive closed-loop control method for the rocket's final stage, the control mode can be detected in real time and switched autonomously. This solves the problems of fault detection and control in rocket final stage deorbiting technology, and enables efficient and safe deorbiting missions.
Patent Information
- Application Number
- CN202611578770.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-10-13
- Publication Date
- 2026-11-17
AI Technical Summary
Existing rocket final stage active deorbiting technology lacks real-time detection methods for various types of faults, cannot adaptively switch control modes in stages, and relies on ground-based telemetry and control, resulting in decreased deorbiting accuracy and safety hazards.
An adaptive closed-loop control method is constructed, which involves fault detection, mode decision-making, online trajectory replanning, and closed-loop execution. This method detects propellant leakage, engine ignition failure, and attitude divergence in real time, autonomously switches control modes, and replans the deorbit trajectory online. Closed-loop control is performed using the onboard computer.
Achieving high-quality deorbiting under fault conditions reduces the safety risks of uncontrolled reentry landing point deviation and residual propellant on orbit, thereby improving deorbiting reliability and mission safety.
Smart Images

Figure CN123111018A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of guidance, navigation and control technology for spacecraft, specifically relating to a fault-tolerant adaptive deorbiting closed-loop control method and system for the final stage of a rocket. Background Technology
[0002] After completing its primary mission, the final stage of a launch vehicle must be actively deorbited to burn up in the atmosphere or fall into a predetermined safe landing zone to prevent it from remaining in orbit for an extended period and becoming space debris or a source of residual explosion. Guidance, navigation, and control (GNC) for active deorbiting is a crucial step in ensuring space debris removal and on-orbit safety.
[0003] There are several existing rocket final stage active deorbit control methods, but the existing rocket final stage active deorbit technology has the following problems and disadvantages: (1) It lacks a composite autonomous real-time fault detection method for multiple types of faults (propellant leakage, engine ignition failure, attitude divergence). Faults can often only be discovered by ground post-failure interpretation and cannot be addressed in a timely manner; (2) After a fault occurs, it lacks the ability to adaptively switch control modes according to the fault type and remaining propulsion capacity. It cannot use attitude control (RCS) or retro-propulsion system to replace the main thruster to complete the deorbit when the attitude control (RCS) or retro-propulsion system is still available. (3) It relies heavily on ground-based telemetry and control or pre-calculation, lacking the ability to replan the trajectory online on the onboard computer based on the remaining propellant and multiple constraints (landing point safety zone, heat flow, overload, propellant depletion). When the flight conditions deviate from the design value due to the fault, it still executes according to the preset parameters, and the deorbit accuracy and success rate are significantly reduced. (4) When the propulsion capacity is seriously insufficient or the attitude diverges, there is a lack of safe decay trajectory selection and enhanced passivation (complete discharge of remaining propellant, depressurization, discharge) treatment, which poses safety hazards such as uncontrollable reentry landing point and residual propellant explosion on the orbit polluting the orbital environment. Summary of the Invention
[0004] The purpose of this invention is to provide a fault-tolerant adaptive deorbiting closed-loop control method and system for the final stage of a rocket. This method enables the final stage of the rocket to autonomously detect faults such as propellant leakage, engine ignition failure, and attitude divergence in real time during active deorbiting. Based on the fault type and remaining propulsion capacity, the system autonomously switches control modes in stages on the onboard computer, replans the deorbit trajectory online, and executes the closed-loop control. This allows the rocket to complete the deorbiting mission with high quality even under fault conditions, reducing the safety risks caused by uncontrolled reentry landing point deviation and residual propellant in orbit, without relying on ground-based telemetry and control.
[0005] According to one objective of the present invention, the present invention provides a fault-tolerant adaptive deorbiting closed-loop control method for the final stage of a rocket, constructing an adaptive closed loop of "fault detection → mode decision → online trajectory replanning → closed-loop execution and navigation fusion", including the following steps: Step S1, Detection: Real-time acquisition of navigation information, propulsion system status and actuator feedback; joint judgment of at least two or all of the faults such as propellant leakage, engine ignition failure and attitude divergence, combining threshold judgment and multi-source consistency test; output fault type, severity level and confidence level. Step S2, Decision: Based on the fault type, severity level and remaining propellant status, determine and switch the current control mode between normal main thruster derailment mode, attitude control multi-pulse derailment mode and safety decay and enhanced passivation mode; Step S3, Replanning: Using the current navigation status, remaining propellant mass, and available thrust model after the failure as input, construct an optimal control problem that includes constraints on the landing safety zone, heat flow, overload, and complete propellant consumption. Solve the problem in real time on the onboard computer and output a new ignition time, thrust profile or pulse sequence, and shutdown conditions. Step S4, Execution: The navigation and attitude control system tracks the ignition time, thrust direction profile or pulse sequence and shutdown conditions to perform derailment in a closed loop, and transmits the actual speed increment and attitude tracking error back to the detection and replanning step in real time to form a rolling closed loop until derailment is completed or the reentry interface is entered.
[0006] Furthermore, the propellant leakage is determined by the absolute value of the difference between the estimated propellant mass loss rate and the theoretical consumption rate exceeding a first preset threshold and lasting for a duration exceeding a first preset duration, or by the tank pressure drop rate exceeding a second preset threshold; the engine ignition failure is determined by the combustion chamber pressure not reaching a third preset threshold or the measured speed increment not reaching a fourth preset threshold within the ignition determination time after commanded ignition; the attitude divergence is determined by the absolute value of the attitude angular velocity exceeding a fifth preset threshold and the attitude error integral continuously increasing, or by the angle between the thrust direction and the desired thrust direction exceeding a sixth preset threshold and lasting for a duration exceeding the first preset threshold.
[0007] Furthermore, the normal main thruster deorbiting mode is performed by the main engine with a single or limited number of deorbiting ignitions; the attitude control multi-pulse deorbiting mode, when the main thruster is unavailable but the attitude control or retro-rocket propulsion system is available, utilizes the attitude control engine (RCS, Reaction Control System) or retro-rocket to perform multi-pulse discrete deorbiting, and optimizes the timing, direction, and amplitude of each pulse online; the safety decay and enhanced passivation mode, when the propulsion capability is severely insufficient, performs enhanced passivation by discharging the remaining propellant, depressurizing, and discharging, and adjusts the attitude to maximize aerodynamic drag or guide into a preset safety decay trajectory, so that the natural reentry landing point is far away from densely populated areas.
[0008] Furthermore, the determination and switching of the current control mode is implemented using a finite state machine, and switching anti-shake criteria and priority rules are set; the switching anti-shake criteria is that the mode switching is triggered only after the fault state is continuously maintained for a preset number of detection cycles or a preset duration; the priority rules are that the safety decay and enhanced passivation modes have higher priority than the attitude control multi-pulse derailment mode, and the attitude control multi-pulse derailment mode has higher priority than the normal main propulsion derailment mode.
[0009] Furthermore, the objective function of the optimal control problem is to minimize propellant consumption or maximize landing safety margin. The constraints also include terminal perigee altitude, attitude constraints, and thrust amplitude and direction limitations. A lightweight convex optimization method (such as sequential convex programming) is adopted and combined with simplified pseudospectral discretization to solve the problem in real time on the onboard computer. The optimized solution is updated periodically through a rolling time-domain strategy.
[0010] Furthermore, steps S1, S2, and S3 are executed in a cycle of 0.1s to 1s, and the real-time solution time is no more than 2s, and no more than 5s in the degraded solution mode.
[0011] Furthermore, the remaining propellant mass is estimated by integrating accelerometer data with tank pressure and temperature models, and the available velocity increment is updated in real time; navigation uses INS / GNSS combined navigation, which degrades to pure inertial navigation or introduces star sensor assistance in case of failure.
[0012] Furthermore, the real-time solution adopts a hot-start initial value, sets an upper limit for the number of iterations and a convergence threshold, and performs convex approximation on the landing point safety zone, heat flow, overload and thrust amplitude constraints to keep each iterative subproblem a convex problem; when the calculation time or number of iterations exceeds the limit, it degenerates into two-body perigee or single-constraint landing point guidance or outputs a deorbit command according to a preset pulse plan.
[0013] Furthermore, the attitude control multi-pulse derailment mode is only used when the remaining available velocity increment meets the minimum derailment capability threshold; when the remaining available velocity increment is insufficient, it switches to the safety decay and enhanced passivation mode.
[0014] The present invention also provides a fault-tolerant adaptive deorbiting closed-loop control system for the final stage of a rocket, including a fault detection unit, a mode decision unit, an online trajectory replanning unit, and a closed-loop execution and navigation fusion unit installed on the onboard computer of the final stage of the launch vehicle; The fault detection unit is used to collect navigation information, propulsion system status and actuator feedback, and to perform a joint judgment combining threshold judgment and multi-source consistency test on at least two or all of the faults such as propellant leakage, engine ignition failure and attitude divergence, and output the fault type, severity level and confidence level. The mode decision unit is connected to the fault detection unit and is used to determine and switch the current control mode between normal main thruster derailment mode, attitude control multi-pulse derailment mode and safety decay and enhanced passivation mode according to the fault type, severity level and remaining propellant status. The online trajectory replanning unit is connected to the mode decision unit and is used to construct an optimal control problem that includes the landing point safety zone, heat flow, overload and propellant complete consumption constraints, with the current navigation state, remaining propellant mass and available thrust model after the fault as input, and solve it in real time, and output a new ignition time, thrust direction profile or pulse sequence and shutdown conditions. The closed-loop execution and navigation fusion unit is connected to the online trajectory replanning unit and is used to drive the navigation and attitude control system to track the ignition time, thrust direction profile or pulse sequence and shutdown conditions to perform closed-loop derailment, and transmit the actual speed increment and attitude tracking error back to the fault detection unit and the online trajectory replanning unit in real time. The rocket's final stage fault-tolerant adaptive deorbiting closed-loop control system implements the rocket's final stage fault-tolerant adaptive deorbiting closed-loop control method described above.
[0015] The technical solution of this invention constructs an autonomous fault-tolerant deorbiting closed-loop control system for rockets by organically linking four stages: fault detection, mode decision-making, online trajectory replanning, and closed-loop execution. Its beneficial effects are as follows: It employs a joint fault determination method combining threshold judgment and multi-source consistency verification, which can simultaneously identify three typical faults: propellant leakage, engine ignition failure, and attitude divergence, and output the fault type, severity level, and confidence level, significantly reducing the probability of misjudgment and missed judgment; it autonomously switches between three deorbiting modes based on the fault type, severity, and remaining propellant status, enabling the rocket to... Even under fault conditions such as leakage, ignition failure, and attitude divergence, the final stage of the rocket can still deorbit, avoiding uncontrolled reentry due to main thruster failure. Online trajectory replanning solves the optimal control problem in real time with constraints such as the landing safety zone, heat flux, overload, and complete propellant consumption. Closed-loop execution feeds back the actual velocity increment and attitude tracking error in real time to form rolling optimization to absorb navigation and execution deviations online. In summary, this invention effectively reduces the safety risks caused by uncontrolled reentry landing deviation and residual propellant in orbit without ground control intervention, and improves the overall reliability and mission safety of final stage deorbiting. Attached Figure Description
[0016] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.
[0017] Figure 1 This is a block diagram of the overall architecture of the rocket final stage fault-tolerant adaptive deorbiting GNC system of the present invention; Figure 2 This is a schematic diagram of the mode switching finite state machine of the present invention; Figure 3 This is a schematic diagram of the online trajectory replanning method of the present invention; Figure 4 This is a schematic diagram showing the simulation comparison curves of the derailment trajectory and landing point under a typical fault scenario (engine ignition failure) of the present invention, with and without fault-tolerant derailment.
[0018] In the diagram: 1. Fault detection module; 2. Mode decision module; 3. Online trajectory replanning module; 4. Closed-loop execution and navigation fusion module; 5. Onboard computer. Detailed Implementation
[0019] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. 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.
[0020] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting this invention.
[0021] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. Furthermore, the terms "installed," "connected," and "linked" should be interpreted broadly; for example, they may refer to a fixed connection, a detachable connection, or an integral connection; they may refer to a mechanical connection or an electrical connection; they may refer to a direct connection or an indirect connection through an intermediate medium; and they may refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.
[0022] Example 1 like Figure 2 and Figure 3 As shown in the figure, this embodiment provides a fault-tolerant adaptive deorbiting closed-loop control method for the final stage of a rocket, including the following steps: Step S1, Detection: Real-time acquisition of navigation information, propulsion system status and actuator feedback; joint judgment of at least two or all of the faults such as propellant leakage, engine ignition failure and attitude divergence, combining threshold judgment and multi-source consistency test; output fault type, severity level and confidence level. Step S2, Decision: Based on the fault type, severity level and remaining propellant status, determine and switch the current control mode between normal main thruster derailment mode, attitude control multi-pulse derailment mode and safety decay and enhanced passivation mode; Step S3, Replanning: Using the current navigation status, remaining propellant mass, and available thrust model after the failure as input, construct an optimal control problem that includes constraints on the landing safety zone, heat flow, overload, and complete propellant consumption. Solve the problem in real time on the onboard computer and output a new ignition time, thrust profile or pulse sequence, and shutdown conditions. Step S4, Execution: The navigation and attitude control system tracks the ignition time, thrust direction profile or pulse sequence and shutdown conditions to perform derailment in a closed loop, and transmits the actual speed increment and attitude tracking error back to the detection and replanning step in real time to form a rolling closed loop until derailment is completed or the reentry interface is entered.
[0023] In this embodiment, after the final stage of the launch vehicle separates from the satellite, it enters the deorbit preparation phase, which is carried out according to the following steps: S101: The system initializes and enters normal monitoring mode; S102: The fault detection module runs in a cycle (0.1s~1s) and executes step S1 to determine in real time whether there is a propellant leak, engine ignition failure or attitude divergence fault. S103: If there is no fault, the nominal off-rail guidance is executed. Existing iterative angle of attack or convex optimization off-rail methods can be reused. The main engine performs one or a limited number of off-rail ignitions in the normal main propulsion off-rail mode. S104: If a fault is detected, proceed to step S2, switching modes according to the fault type and remaining propellant status at a cycle of 0.1s to 1s: for minor leakage, adjust the maximum ignition duration and angle of attack and continue main thrust; for ignition failure or moderate fault and when the attitude control or reverse thrust system is available, switch to attitude control multi-pulse mode and call the online planning module to generate a pulse sequence; for severe faults such as severely insufficient propulsion capacity or attitude divergence, switch to safety decay and enhanced passivation mode, discharge remaining propellant, depressurize, discharge, and adjust attitude. In this embodiment, the determination and switching of the above modes are implemented using a finite state machine, and switching anti-shake criteria and priority rules are set: the switching anti-shake criteria is that the mode switching is triggered only after the fault state is continuously maintained for a preset number of detection cycles (such as 3 consecutive detection cycles) or a preset duration, so as to prevent frequent mode oscillation switching; the priority rule is that the priority of the safety decay and enhanced passivation mode is higher than the attitude control multi-pulse derailment mode, the priority of the attitude control multi-pulse derailment mode is higher than the normal main propulsion derailment mode, and the current mode is switched to the high priority mode when the entry conditions of the high priority mode are met.
[0024] S105: After the replanning generation instruction is executed, steps S3 and S4 are performed. The automatic driving / attitude control system performs closed-loop tracking until derailment is completed or the reentry interface is entered. S106: Records telemetry data throughout the entire process for post-event analysis and evaluation on the ground.
[0025] Example 2 In this embodiment, the specific method for fault detection includes: Propellant mass estimation is performed using the following formula: ; in, For the initial propellant mass, This is the theoretical (nominal) propellant consumption rate. It was obtained by correcting the tank pressure and temperature model.
[0026] Propellant leakage criterion: When the absolute value of the difference between the estimated propellant mass degradation rate and the theoretical consumption rate satisfies Or the rate of pressure drop in the tank exceeds And lasting longer than At a certain time, a propellant leak is determined, and the leak rate can be estimated by the statistical average of the differences.
[0027] Engine ignition failure criteria: after the ignition command is issued Combustion chamber pressure within a time period or measured speed increment If the ignition fails, it is determined that the ignition has failed.
[0028] Attitude divergence criterion: when Furthermore, the attitude error integral exceeds the limit, or the angle between the thrust direction and the desired thrust direction exceeds the limit. Exceeding the sixth preset threshold And continue for the first preset duration At that time, the attitude divergence is determined. Among them... For attitude angular velocity, This is the angular velocity threshold.
[0029] In this embodiment, the joint determination combining threshold judgment and multi-source consistency verification is performed as follows: The information used for fault detection is divided into three independent information sources. The first category is navigation sources, including position, velocity, attitude, angular velocity, and apparent acceleration output by inertial navigation, position and velocity output by satellite navigation, and attitude output by star sensors. The second category is propulsion system sources, including tank pressure, tank temperature, liquid level or propellant book mass, engine combustion chamber pressure, and valve status. The third category is actuator feedback sources, including engine ignition commands, servo swing angle or attitude control nozzle commands, and measured velocity increments. The same physical quantity is observed or calculated by two or more independent information sources, and the results of each information source are cross-validated pairwise. The corresponding fault is confirmed only when the threshold judgment result is mutually corroborated by two or more independent information sources. If only a single information source exceeds the limit and the other information sources do not corroborate it, it is set as a suspected fault and continues to be observed without immediately triggering mode switching.
[0030] The multi-source consistency check includes the following cross-verification relationships: (1) Propellant consumption consistency check, the actual consumption inverted by the apparent acceleration integral, the consumption calculated by the bookkeeping formula, and the consumption inverted by the tank pressure-temperature model are compared pairwise and the normalized residuals are calculated. When any two residuals exceed the consistency threshold at the same time and continue for a preset verification time, it is determined that the propellant leakage has been confirmed by multiple sources. If only the pressure-temperature model indicates abnormality and the accelerometer integral consumption is normal, it is determined that the pressure or temperature channel is suspicious and the leakage is not confirmed; (2) Thrust and ignition status consistency check, the actual thrust identified by the apparent acceleration, the thrust converted by the combustion chamber pressure, and the expected thrust corresponding to the ignition command are compared. When the ignition command has been issued and the two independent information sources of combustion chamber pressure and apparent acceleration are at the ignition determination time, the propellant leakage is confirmed by multiple sources. Ignition failure is confirmed only when the internal pressure does not reach the corresponding threshold. If only the combustion chamber pressure is low and the acceleration response is normal, the pressure measurement channel is suspected and ignition failure is not confirmed. (3) Attitude consistency verification: the attitude and angular velocity output by the inertial navigation are compared with the attitude output by the star sensor. When the inertial navigation indicated angular velocity exceeds the limit and the attitude error integral continues to increase and is consistent with the attitude residual of the star sensor exceeding the limit, attitude divergence is confirmed. If only the inertial navigation channel exceeds the limit and the attitude of the star sensor is normal, the inertial device is suspected and the navigation source is switched and attitude divergence is not confirmed. The above normalized residual adopts threshold test. After the residual is normalized by the measurement noise variance, it is compared with the consistency threshold. If the preset number of detection cycles exceeds the limit, the corresponding information source is determined to be inconsistent. Suspicious information sources are isolated and the weighted result of the other consistent information sources is used as the standard.
[0031] In this embodiment, the joint judgment is output by combining threshold judgment and multi-source voting: when two or more independent information sources exceed the corresponding fault threshold at the same time and are consistent with each other, the fault is confirmed and the confidence level is set to high; when only one information source exceeds the limit, a suspected fault is output, the confidence level is set to low and the current mode is maintained for continued observation; when the conclusions of each information source are contradictory and cannot be arbitrated, the information sources are accepted according to the preset information source priority and the confidence level is set to medium.
[0032] In this embodiment, the severity level is divided into three levels: minor, moderate, and severe, which are determined by the degree of impact of the fault on propulsion and attitude control capabilities, as well as the margin of the remaining available speed increment relative to the speed increment required for deorbiting.
[0033] In this embodiment, the remaining propellant mass is estimated by integrating accelerometer data with a tank pressure and temperature model: the actual consumption is obtained by integrating the actual acceleration measured by the accelerometer during engine operation, and then weighted and fused with the consumption estimated by the tank pressure and temperature model to update the remaining propellant mass and available velocity increment in real time. ΔV This serves as input for mode decision-making and online trajectory replanning. Navigation employs a combination of INS / GNSS navigation. When GNSS signal interference or interruption is detected, navigation is downgraded to pure inertial navigation or a star sensor is introduced for assistance to ensure navigation accuracy under fault conditions.
[0034] Example 3 This embodiment of attitude control multi-pulse online planning includes: When switching to the attitude control multi-pulse deorbiting mode, the continuous optimal control problem is transformed into a discrete pulse optimization problem: the decision variables are the time, direction, and amplitude of each pulse (or the velocity increment vector corresponding to each pulse), and the time, direction, and amplitude of each pulse are optimized online; the objective function is to minimize propellant consumption or maximize the landing safety margin, or the lowest terminal perigee or the smallest landing deviation can be selected; constraints include terminal perigee altitude, upper limit of pulse count (e.g., no more than 10 times), single pulse capability limit, attitude reachability range, and landing safety zone; sequential convex programming (SCvx) or mixed integer programming is used for approximate solution, and the computational load is reduced by limiting the number of pulses, and the pulse ignition sequence is generated online in real time on the onboard computer; the solution results are updated in a rolling time domain strategy with a period of 0.1s to 1s to absorb navigation errors and execution deviations.
[0035] In this embodiment, the constraints of the optimal control problem also include heat flux constraints, overload constraints, and propellant complete consumption constraints: the heat flux constraint limits the stagnation heat flux at each point on the reentry trajectory to below the maximum heat flux that the thermal protection system can withstand; the overload constraint limits the axial and normal overloads during flight to below the maximum overload allowed by the structure. Both are applied point-by-point along the reentry trajectory and convex to a second-order cone or a first-order approximation during the solution process; the propellant complete consumption constraint requires that the remaining propellant mass at the end of deorbit ignition does not exceed a set residual upper limit, so that deorbit ignition consumes as much available propellant as possible. When the propellant cannot be consumed due to pulse capability limitations, a safe decay and enhanced passivation mode is entered to discharge the remaining propellant. These constraints, together with the landing safety zone, terminal perigee altitude, attitude reachability range, and single-pulse capability limitations, constitute the constraint set of the optimal control problem.
[0036] In this embodiment, the shutdown conditions are determined in real time based on the online planning results: when the cumulative applied velocity increment reaches the command velocity increment of the current ignition or the pulse sequence, and the remaining ignition duration reaches the planned value, or when the remaining propellant mass is detected to drop to the shutdown threshold or the combustion chamber pressure is abnormal, a shutdown command is output to end the current ignition; after each shutdown, the remaining propellant and available velocity increment are re-estimated, and subsequent pulses or whether to switch modes are determined accordingly.
[0037] Example 4 The safety attenuation and enhanced passivation modes in this embodiment include: When switching to the safe decay and enhanced passivation mode: calculate the current orbital natural decay time and the predicted envelope of the landing point; select an attitude to maximize aerodynamic drag, or guide the stage into the corresponding South Pacific safe zone decay corridor calculated online or selected by looking up a table based on the current orbital elements and atmospheric density model; simultaneously perform enhanced passivation treatment, completely discharge the remaining propellant, depressurize and discharge it, so that the final stage lands far away from densely populated areas during uncontrolled natural reentry, and there is no longer a risk of residual propellant exploding in orbit; and retain a certain amount of propellant for a final attitude correction when conditions permit.
[0038] Example 5 The real-time engineering measures for online replanning in this embodiment include: Based on the attitude control multi-pulse online planning described in Example 3, the following engineering measures are further adopted to meet the real-time requirements of the rocket: a hot start method is used (the feasible solution or nominal deorbit trajectory of the previous rolling cycle is used as the initial value of SCvx iteration) and the dimensionality of the model is reduced by combining simplified pseudospectral discretization and state and control normalization; nonlinear constraints such as heat flow, overload, and thrust amplitude are convexized into second-order cones or first-order approximations, and the landing safety zone is convexized into a polytope to ensure that each iterative subproblem is a convex problem; an upper limit for the number of iterations (e.g., 10 to 20 times) and a convergence threshold are set, and cone relaxation is introduced to ensure the feasibility and monotonic convergence of the iteration; when the calculation time or the number of iterations exceeds the limit, it degenerates into two-body perigee / landing point single-constraint guidance or a preset pulse plan, so that under extreme fault conditions, the conventional solution can output an executable deorbit command within 2 seconds, and the degraded solution mode does not exceed 5 seconds. The solver configuration corresponding to the 2s upper limit of the solution time is as follows: under the initial value of hot start, it usually converges in 3 to 5 iterations, and the solution time of a single iteration is less than 0.4s. The upper limit of 10 iterations corresponds to a time budget of 4s, leaving a 2s margin; in extreme failure cases, the time can be relaxed to 5s by degrading the solution.
[0039] Combination Figure 4 In a typical fault simulation scenario, the engine ignition fails during deorbiting ignition of the final stage. Without the fault-tolerant control of this invention, the final stage maintains uncontrolled flight on its original trajectory and re-enters naturally, resulting in a large landing point dispersion range and a landing point center deviating from the predetermined safe landing zone. With the fault-tolerant control of this invention, fault detection confirms ignition failure within the ignition decision time, the mode decision switches to the attitude control multi-pulse deorbiting mode, online replanning generates a multi-pulse sequence and executes it in a closed loop, causing the re-entry landing point to converge within the envelope of the predetermined safe landing zone, and the landing point dispersion range is significantly reduced. The simulation results show that under the condition of engine ignition failure, this invention can still control the uncontrolled re-entry landing point deviation within a safe range, verifying the effectiveness of the fault-tolerant deorbiting closed loop; the specific landing point deviation and dispersion range vary depending on the orbital altitude, remaining propulsion capacity, and safe landing zone settings.
[0040] Example 6 The applicable conditions and joint fallback for the attitude control multi-pulse mode in this embodiment include: Attitude control multi-pulse deorbit mode only applies to the remaining available velocity increment. ΔV The primary deorbiting method is employed only when the minimum deorbiting or reentry capability threshold is met. When the remaining propellant is insufficient to meet this threshold, the system directly switches to a safe decay and enhanced passivation mode. This involves releasing the remaining propellant, depressurizing, discharging, and adjusting the attitude to maximize aerodynamic drag or enter a preset safe decay trajectory. This ensures controllable landing and safe on-orbit handling, avoiding reliance solely on attitude control (RCS) for deorbiting due to insufficient capability. The minimum deorbiting or reentry capability threshold is calculated online based on the target's perigee altitude and current orbital energy.
[0041] Example 7 like Figure 1 As shown in the figure, this embodiment presents a fault-tolerant adaptive deorbiting closed-loop control system for the final stage of a rocket, installed in the onboard computer 5 of the final stage of the launch vehicle. The fault detection unit, mode decision unit, online trajectory replanning unit, and closed-loop execution and navigation fusion unit are respectively implemented by the fault detection module 1, mode decision module 2, online trajectory replanning module 3, and closed-loop execution and navigation fusion module 4 running in the onboard computer 5.
[0042] The fault detection module 1 is connected to the sensors of the navigation / attitude control system and the propulsion system respectively, and collects navigation information (position, speed, attitude, angular velocity), propulsion system status (propellant mass estimate, tank pressure, engine combustion chamber pressure, valve status) and actuator feedback. It performs a joint judgment combining threshold judgment and multi-source consistency test on at least two or all of the faults such as propellant leakage, engine ignition failure and attitude divergence, and outputs the fault type, severity level and confidence level to the mode decision module 2. The mode decision module 2 determines and switches the current control mode between normal main thruster derailment mode, attitude control multi-pulse derailment mode, and safety decay and enhanced passivation mode based on the fault type, severity level, and remaining propellant status using a finite state machine. It also sets switching anti-jitter criteria and priority rules: mode switching is only triggered after the fault state is maintained for a preset number of detection cycles (e.g., 3 consecutive detection cycles) or a preset duration to prevent frequent mode oscillation switching; the priority of the safety decay and enhanced passivation mode is higher than that of the attitude control multi-pulse derailment mode, and the priority of the attitude control multi-pulse derailment mode is higher than that of the normal main thruster derailment mode. The online trajectory replanning module 3 takes the current navigation status, remaining propellant mass, and available thrust model after the fault as input to construct an optimal control problem that includes the landing safety zone, heat flow, overload, and complete propellant consumption constraints. It uses SCvx and combines simplified pseudospectral discretization to solve the problem in real time on the onboard computer 5, and outputs a new ignition time, thrust direction profile or pulse sequence, and shutdown conditions. The closed-loop execution and navigation fusion module 4 drives the navigation and attitude control system to track the ignition time, thrust direction profile or pulse sequence and shutdown conditions to perform closed-loop derailment, and transmits the actual speed increment and attitude tracking error back to the fault detection module 1 and the online trajectory replanning module 3 in real time to form a rolling closed loop.
[0043] The fault-tolerant adaptive deorbiting closed-loop control system for rocket final stage faults described in this embodiment realizes the fault-tolerant adaptive deorbiting closed-loop control method for rocket final stage faults as described in any one of claims 1 to 9.
[0044] All of the above modules are software modules running on the rocket's onboard computer 5. They can be used as enhancement modules and compatible with existing deorbiting methods (iterative angle of attack, ground injection) without significantly increasing hardware.
[0045] In the above embodiments: the solution method for the optimal control problem is not limited to SCvx, but can also be achieved using convex optimization methods, model predictive control methods, or quadratic programming methods; the fault types are not limited to propellant leakage, engine ignition failure, and attitude divergence, but can also be extended to include tank overpressure, pipeline blockage, actuator jamming, and other faults; the determination of attitude divergence is not limited to the integral criterion of angular velocity and attitude error, but can also use the criterion of exceeding the limit of the angle between the thrust direction and the desired thrust direction; the selection of the safe decay trajectory is not limited to online calculation, but can also be achieved by looking up a table based on the atmospheric density model; the navigation method is not limited to INS / GNSS combined navigation, but can also be achieved by inertial navigation and star sensor combined navigation.
[0046] Those skilled in the art will understand that the specific parameters (detection period, threshold) in the above embodiments are as follows: δ 1. T 1. δ 2. t ign , P th , ΔV th , ω th The values (such as the upper limit of the number of pulses) are all exemplary values and can be adjusted according to the propulsion system characteristics, attitude control capabilities and mission safety requirements of different rocket final stages. All equivalent transformations, substitutions and combinations made based on the concept of this invention are within the scope of protection claimed by this invention.
[0047] Compared with the prior art, the present invention has the following beneficial effects: (1) This invention significantly improves the success rate of launch vehicle final stage deorbiting missions under fault conditions such as propellant leakage, engine ignition failure, and attitude divergence, reduces the safety risks caused by uncontrolled reentry and on-orbit residual propellant, and improves the safety and compliance of active space debris removal and orbital environmental protection. (2) This invention achieves fully / highly autonomous operation throughout the entire process. Fault detection, mode switching and trajectory replanning are all completed on the onboard computer, without relying on ground measurement and control, and adapting to extreme scenarios such as communication interruption; (3) This invention maximizes the use of residual propulsion capability and attitude control / reverse thrust through multi-level mode switching and online trajectory replanning, while taking into account landing safety, propellant depletion and deorbit accuracy. (4) The online replanning of this invention adopts lightweight convex optimization / SCvx, and the computational load is controllable (the conventional solution does not exceed 2s and the degraded solution does not exceed 5s), which is suitable for the limited onboard computing resources of commercial rockets; (5) This invention can be implemented independently as described in this invention, or it can be compatible with existing derailment methods (iterative angle of attack, ground injection) and used in combination in software form. It has low engineering implementation cost, good scalability, and can be further extended to be compatible with the subsequent precise derailment design of the track-detaining platform.
[0048] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.
Claims
1. A rocket terminal stage fault-tolerant adaptive orbit departure closed-loop control method, characterized in that, Includes the following steps: Step S1, Detection: Real-time acquisition of navigation information, propulsion system status and actuator feedback; joint judgment of at least two or all of the propellant leakage, engine ignition failure and attitude divergence faults by combining threshold judgment and multi-source consistency test; output fault type, severity level and confidence level. Step S2, Decision: Based on the fault type, severity level and remaining propellant status, determine and switch the current control mode between normal main thruster derailment mode, attitude control multi-pulse derailment mode and safety decay and enhanced passivation mode; Step S3, Replanning: Using the current navigation status, remaining propellant mass, and available thrust model after the failure as input, construct an optimal control problem that includes constraints on the landing safety zone, heat flow, overload, and complete propellant consumption. Solve the problem in real time on the onboard computer and output a new ignition time, thrust profile or pulse sequence, and shutdown conditions. Step S4, Execution: The navigation and attitude control system tracks the ignition time, thrust direction profile or pulse sequence and shutdown conditions to perform derailment in a closed loop, and transmits the actual speed increment and attitude tracking error back to the detection and replanning step in real time to form a rolling closed loop until derailment is completed or the reentry interface is entered.
2. The rocket final stage fault-tolerant adaptive deorbiting closed-loop control method according to claim 1, characterized in that, The propellant leakage is determined when the absolute value of the difference between the estimated propellant mass loss rate and the theoretical consumption rate exceeds a first preset threshold and lasts for a duration exceeding a first preset duration, or when the tank pressure drop rate exceeds a second preset threshold; the engine ignition failure is determined when the combustion chamber pressure does not reach a third preset threshold or the measured speed increment does not reach a fourth preset threshold within the ignition determination time after commanded ignition; the attitude divergence is determined when the absolute value of the attitude angular velocity exceeds a fifth preset threshold and the attitude error integral continues to increase, or when the angle between the thrust direction and the desired thrust direction exceeds a sixth preset threshold and lasts for a duration exceeding the first preset threshold.
3. The rocket final stage fault-tolerant adaptive deorbiting closed-loop control method according to claim 1, characterized in that, The normal main thruster deorbiting mode involves the main engine performing a single or limited number of deorbiting ignitions. The attitude control multi-pulse deorbiting mode, when the main thruster is unavailable but the attitude control or retro-rocket propulsion system is available, utilizes the attitude control engine or retro-rocket to perform multi-pulse discrete deorbiting and optimizes the timing, direction, and amplitude of each pulse online. The safety decay and enhanced passivation mode, when propulsion capability is severely insufficient, performs enhanced passivation by discharging remaining propellant, depressurizing, and discharging, and adjusts the attitude to maximize aerodynamic drag or guide the vehicle into a preset safety decay trajectory, ensuring that the natural reentry landing point is far away from densely populated areas.
4. The rocket final stage fault-tolerant adaptive deorbiting closed-loop control method according to claim 1, characterized in that, The determination and switching of the current control mode is implemented using a finite state machine, and switching anti-shake criteria and priority rules are set; the switching anti-shake criteria is that the mode switching is triggered only after the fault state is continuously maintained for a preset number of detection cycles or a preset duration. The priority rule is that the safety decay and enhanced passivation mode has a higher priority than the attitude control multi-pulse deorbiting mode, and the attitude control multi-pulse deorbiting mode has a higher priority than the normal main propulsion deorbiting mode.
5. The rocket final stage fault-tolerant adaptive deorbiting closed-loop control method according to claim 1, characterized in that, The objective function of the optimal control problem is to minimize propellant consumption or maximize landing safety margin. The constraints also include terminal perigee altitude, attitude constraints, and thrust amplitude and direction limitations. A lightweight convex optimization method combined with simplified pseudospectral discretization is used to solve the problem in real time on the onboard computer, and the optimized solution is updated periodically through a rolling time-domain strategy.
6. The rocket final stage fault-tolerant adaptive deorbiting closed-loop control method according to claim 1, characterized in that, Steps S1, S2 and S3 are run in a cycle of 0.1s to 1s. The real-time solution time is no more than 2s and no more than 5s in the degraded solution mode.
7. The rocket final stage fault-tolerant adaptive deorbiting closed-loop control method according to claim 1, characterized in that, The remaining propellant mass is estimated by integrating accelerometer data with tank pressure and temperature models, and the available velocity increment is updated in real time. Navigation uses INS / GNSS combined navigation, which is downgraded to pure inertial navigation or assisted by star sensors in case of failure.
8. The rocket final stage fault-tolerant adaptive deorbiting closed-loop control method according to claim 1, characterized in that, The real-time solution uses hot-start initial values, sets an upper limit for the number of iterations and a convergence threshold, and performs convex approximation on the landing point safety zone, heat flow, overload and thrust amplitude constraints to keep each iterative subproblem a convex problem; when the calculation time or number of iterations exceeds the limit, it degenerates into two-body perigee or single-constraint landing point guidance or outputs deorbit command according to a preset pulse plan.
9. The rocket final stage fault-tolerant adaptive deorbiting closed-loop control method according to claim 1, characterized in that, The attitude control multi-pulse derailment mode is only used when the remaining available velocity increment meets the minimum derailment capability threshold; when the remaining available velocity increment is insufficient, it switches to the safety decay and enhanced passivation mode.
10. A fault-tolerant adaptive deorbiting closed-loop control system for the final stage of a rocket, characterized in that, This includes a fault detection unit, a mode decision unit, an online trajectory replanning unit, and a closed-loop execution and navigation fusion unit installed on the onboard computer of the final stage of the launch vehicle; The fault detection unit is used to collect navigation information, propulsion system status and actuator feedback, and to perform a joint judgment combining threshold judgment and multi-source consistency test on at least two or all of the faults of propellant leakage, engine ignition failure and attitude divergence, and output the fault type, severity level and confidence level. The mode decision unit is connected to the fault detection unit and is used to determine and switch the current control mode between normal main thruster derailment mode, attitude control multi-pulse derailment mode and safety decay and enhanced passivation mode according to the fault type, severity level and remaining propellant status. The online trajectory replanning unit is connected to the mode decision unit and is used to construct an optimal control problem that includes the landing point safety zone, heat flow, overload and propellant complete consumption constraints, with the current navigation state, remaining propellant mass and available thrust model after the fault as input, and solve it in real time, and output a new ignition time, thrust direction profile or pulse sequence and shutdown conditions. The closed-loop execution and navigation fusion unit is connected to the online trajectory replanning unit and is used to drive the navigation and attitude control system to track the ignition time, thrust direction profile or pulse sequence and shutdown conditions to perform closed-loop derailment, and transmit the actual speed increment and attitude tracking error back to the fault detection unit and the online trajectory replanning unit in real time. The rocket's final stage fault-tolerant adaptive deorbiting closed-loop control system implements the rocket's final stage fault-tolerant adaptive deorbiting closed-loop control method as described in any one of claims 1 to 9.
