Rocket electrical system control method and device

Abstract

The invention belongs to the technical field of electrical control, and discloses a rocket electrical system control method and device. The method comprises the following steps: collecting flight state and external environment information of a rocket; the flight state information is analyzed through a data processing unit, and potential abnormal conditions are identified; on the basis of identifying potential abnormal conditions, a dynamic model of the rocket electrical system is established, and an optimal control strategy of rocket electrical system parameters is formed by utilizing a dynamic model predictive control MPC algorithm according to the flight state and environment data updated in real time; a feedback mechanism based on an error trend model is established, and a gain matrix K is adjusted in real time, so that an optimization control strategy adapts to the change of internal parameters and the real-time change of an external environment; through regular simulation and actual flight testing, the stability of a feedback mechanism based on an error trend model under a fault condition is evaluated and adjusted. According to the invention, the safety of the spacecraft and ground personnel is greatly improved.

Description

Technical Field

[0001] The present invention belongs to the field of electrical control technology, and in particular relates to a rocket electrical system control method and device. Background Art

[0002] With the development of science and technology and the increase in application needs, competition in the field of space exploration and technology is becoming increasingly fierce. For spacecraft, reliability, safety and efficiency are one of the key factors, especially the power control system of the launch vehicle has extremely high requirements, because the rocket needs to cope with various extreme challenges such as severe weather and ensure the ability to operate safely and stably.

[0003] Traditional rocket electrical system control is usually implemented using fixed-configuration controllers and rigid control strategies. This solution makes the rocket prone to failure or failure to meet the expected precise control requirements when faced with complex nonlinear systems such as the rocket's electrical system, power system, thermal control system, and unknown disturbances. In addition, it is difficult to make corresponding adjustments and optimizations for different models of rockets and the environmental conditions in which they are located, which affects their working performance and service life.

[0004] Existing technical solutions have the following problems: traditional control logic and methods lack the ability to adapt to changing environmental conditions; they rely on fixed parameter settings, making it difficult to fully tap the potential of the rocket electrical system and maximize resource utilization; and insufficient emergency response to emergencies may lead to uncontrollable or even paralyzed systems. Summary of the invention

[0005] In order to overcome the problems existing in the related art, the disclosed embodiments of the present invention provide a method and device for controlling a rocket electrical system.

[0006] The technical solution is as follows: a rocket electrical system control method, the method comprising the following steps:

[0007] S1, collects the rocket’s flight status and external environment information;

[0008] S2, analyzes the flight status and external environment information through the data processing unit to identify potential abnormal situations;

[0009] S3, based on the identification of potential abnormal conditions, a dynamic model of the rocket electrical system is established, and the dynamic model predictive control MPC algorithm is used to form an optimization control strategy for the rocket electrical system parameters according to the real-time updated flight status and environmental data;

[0010] S4, establish a feedback mechanism based on the error trend model to adjust the gain matrix K in real time so that the optimization control strategy can adapt to the real-time changes of internal parameters and external environment;

[0011] S5, evaluate and adjust the stability of the feedback mechanism based on the error trend model under fault conditions through regular simulation and actual flight tests.

[0012] In step S1, the flight status is collected in real time by sensors and measuring equipment on the rocket and transmitted to the controller. The flight status includes: position, altitude, speed, acceleration, temperature, attitude and pressure;

[0013] The external environmental information is acquired by sensors and environmental detection equipment on the rocket and transmitted to the controller. The external environmental information includes factors such as meteorological conditions, terrain features, and geological structures that affect the aircraft and routes; specifically including: air pressure, temperature, wind speed, atmospheric density, radiation levels, and magnetic field strength.

[0014] In step S2, the data processing unit pre-processes the flight status information collected by the sensor, predicts the future flight status based on the historical data and physical model of the rocket's flight status, uses the dynamic model predictive control algorithm to make short-term predictions on the position, speed and attitude, and compares the flight status data with the preset safety threshold. When a parameter exceeds the safety range, it is marked as a potential abnormality;

[0015] The dynamic model predictive control algorithm calculates the control input using the following formula:

[0016] μ(t)=K×(x ref -x(t))+Δμ

[0017] Where μ(t) is the control input at time t, K is the gain matrix, and x ref is the reference trajectory state, x(t) is the current state, Δμ is the increment obtained by error adjustment, and Δμ is used to cope with real-time changes in the external environment and internal state;

[0018] The control input is input to the rocket power management module, thermal management module, attitude control module and communication and navigation module; the errors are state error, prediction error and control input smoothing error, the state error is the difference between the current state of the system and the target state, the prediction error is the difference between the expected system state and the reference trajectory in the future prediction window of the dynamic model predictive control MPC algorithm, and the control input smoothing error is the change between the control input of the previous time step and the increment of the current time step, which responds to real-time changes in the external environment and internal state;

[0019] The rocket power management module, thermal management module, attitude control module and communication and navigation module use unified input, take the weighted average of the state error, prediction error and control input smooth error at time t, and subtract the weighted average of the state error, prediction error and control input smooth error at the previous time, i.e., time t-1, to obtain the error increment, i.e., the increment obtained according to the error adjustment.

[0020] Furthermore, the dynamic model predictive control algorithm obtains the optimal solution by the least square method, comprising the following steps:

[0021] The first step is to obtain the state signal at each time point in the future prediction time window;

[0022] The second step is to construct the control input formula using linear regression or nonlinear mapping methods to obtain the state-error relationship function at all future time points;

[0023] The third step is to estimate the state error and the control input variation coefficient to obtain the minimum value of the cost function;

[0024] The fourth step is to find the optimal solution based on the minimum value and obtain the optimal control amount;

[0025] The fifth step is to update the current state according to the optimal solution and determine the new control input.

[0026] In the third step, the cost function consists of a weighted sum of the state error and the control input change, expressed as:

[0027]

[0028] Where J is the weighted sum of the cost function, x(i|t) is the state from time i to time t, T is the order, μ(i|t) is the control input from time i to time t, μ(i-1|t) is the control input from the time before time i to time t, N is the length of the prediction window, and Q and R are weighted matrices used to balance prediction accuracy and control smoothness.

[0029] In step S3, the optimization control strategy includes:

[0030] Under the premise of flight stability, the dynamic model predictive control algorithm predicts energy demand and adjusts electrical system parameters to minimize energy consumption;

[0031] The time response to the external disturbance is smoothed by selecting the weighting matrices Q and R of the cost function, so that the response time is adapted to the external disturbance;

[0032] Setting an error ratio or a continuous time interval safety constraint can prevent electrical system failures and ensure flight safety; in preventing electrical system failures, the performance data of the electrical system is analyzed in real time, and possible system failures are identified based on a predetermined dynamic model predictive control algorithm, and the system automatically switches to an emergency backup plan or notifies the ground control center; when an abnormal situation occurs, an alarm signal is sent in a timely manner, and the remote control terminal intervenes and takes emergency measures for the corresponding components.

[0033] In step S4, the feedback mechanism based on the error trend model is established to adjust the gain matrix K in real time, including:

[0034] The feedback mechanism based on the error trend model collects data in real time and calculates the error, analyzes the error, and identifies the trend. If the error increases significantly and the system deviation is large, the relevant value of K is increased to increase the control input, and the updated value of the gain matrix K is calculated using the aforementioned cost function. The updated gain matrix K is used to calculate the new control input μ(t) and send it to the controller of the rocket electrical system.

[0035] In step S5, the stability of the feedback mechanism based on the error trend model under fault conditions is evaluated and adjusted, including: regularly simulating flight tests to obtain flight status data, control response data, environmental change data and system error data, and actually performing flight tests to obtain actual flight status data, environmental data, error change data and fault information; comparing the data obtained from the simulated flight test with the preset reference trajectory, analyzing the response effects of different optimization control strategies, establishing an error trend model using the error data, analyzing the error change law, adjusting the gain, feedback frequency and response speed parameters of the optimization control strategy according to the error trend model, and testing the influence of each parameter on the error convergence speed, and the simulation test results are used to adjust and optimize the control parameters such as the initial value of the gain matrix K in the feedback mechanism based on the error trend model; using the data obtained from the actual flight test to determine the adaptability of the existing feedback mechanism based on the error trend model using the error change trend over time, comparing the simulated data with the actual flight data, analyzing whether the parameters in the simulation need to be adjusted, evaluating the reliability of the feedback mechanism based on the error trend model under fault conditions through the analysis of the fault data, and optimizing the trigger threshold and response time of the emergency optimization control strategy according to the error feedback results of the actual flight.

[0036] Furthermore, the error data is used to establish an error trend model, including:

[0037] Step 1, quantitative calculation of the incremental value obtained by error adjustment for single factors of flight status data, environmental data, error change data and fault information, comprehensive consideration of the values ​​of three dimensions of simulated state error weight, prediction error weight and control input smooth error, assign a real value between 0 and 1 to each single factor simulation; setting of state error weight, dividing the state error weight into 10 steps according to the fault severity ranking, the final score of the state error weight is recorded as G1; state error prediction error weight, the type of state error prediction error is automatically determined by state marking, the state error prediction error weight is recorded as G2; control input smooth error weight, reasonably assigning its smoothing coefficient according to the simulated control input smooth error;

[0038] Step 2, the error trend exponential smoothing linear line calculation of the fault-related simulation, by calculating the deviation value PAI, the difference linear mean XYU and the high and low side distribution GDBF histogram of the incremental value obtained according to the error adjustment in different time periods, and calculating the error trend exponential smoothing moving linear line indicator HAD; HAD Mt and HAD Nt They are respectively the exponentially smoothed moving linear average of the error trend in the M time interval and the exponentially smoothed moving linear average of the error trend in the N time interval at the tth moment, where M>N;

[0039] (1) Calculate the HAD value using an exponentially decreasing weighted moving average. The expression is:

[0040]

[0041] In the formula, HAD Nt-1 is the exponentially smoothed moving linear average of the error trend in N time intervals at time t-1, N is the length of the prediction window, and A t is the exponential smoothing moving linear constant at the tth moment, M is the Mth time interval, HAD Mt-1 (M-1) is the exponentially smoothed moving linear average of the error trend in the M time interval at time t-1;

[0042] (2) Calculate the PAI value, the HAD value of the N time interval minus the HAD value of the M time interval. When the error spreads more, the HAD of the N time interval is higher than the HAD of the M time interval. The expression is:

[0043] PAI t =HAD Nt -HAD Mt

[0044] Where PAI t is the PAI value at time t;

[0045] (3) Calculate the XYU value, which is the time interval error trend exponential smoothing moving linear line of the PAI line. The expression is:

[0046]

[0047] In the formula, XYU pt is the XYU value at time t, XYU pt-1 is the XYU value at time t-1, and p is a time different from time t;

[0048] (4) Calculate the moving smoothing trend value, and let the moving smoothing trend value represent the overall development trend of the error. The current trend value is calculated using the high and low edge distribution, and the expression is:

[0049] GDBF t =2(PAI t -XYU pt )

[0050] In the formula, GDBF t is the moving smoothing trend value at time t;

[0051] A crossover signal occurs when the shorter-period HAD crosses the longer-period HAD.

[0052] Step 3, visualization of error trend prediction, after calculating the incremental value obtained by error adjustment and the PAI, XYU and PAIF values ​​in the specified time period and time cycle of the single factor impact condition simulation of flight status data, environmental data, error change data and fault information, use the visual window visualization framework to draw curves according to time periods with different weights, and highlight the intersection points of fast and slow lines and the points crossing the horizontal axis. By providing the overall change trend of multiple curves and the distribution of key points, it helps users quickly grasp the development and evolution of the entire error and the potential change trend in the future;

[0053] By automatically extracting the simulation links of the sliding window where the simulation is located; automatically sorting the links of the simulation type in the error source code; let T represent the number of all error links, linanf c represents the links sorted as c, then the fth c Importance calculation of links:

[0054]

[0055] In the formula, G(lianf c ) is the f c The important value of a link, c is the sequence number;

[0056] Change the link importance to a value between 50 and 100 as the control input smoothing error weight, denoted as G3, and the expression is:

[0057] G3=sigmoid(G(lianf c ))×100.

[0058] Another object of the present invention is to provide a rocket electrical system control device, which implements the above-mentioned rocket electrical system control method, and the device comprises:

[0059] Information collection module, used to collect the rocket's flight status and external environment information;

[0060] A potential abnormal situation identification module is used to analyze the flight status and external environment information through a data processing unit to identify potential abnormal situations;

[0061] The dynamic model predictive control algorithm prediction module is used to establish a dynamic model of the rocket electrical system based on the identification of potential abnormal conditions, and use the dynamic model predictive control MPC algorithm to form an optimization control strategy for the rocket electrical system parameters according to the real-time updated flight status and environmental data;

[0062] Feedback mechanism establishment module, used to establish a feedback mechanism based on the error trend model, adjust the gain matrix K in real time, and make the optimization control strategy adapt to the real-time changes of internal parameters and external environment;

[0063] An evaluation and adjustment module is used to evaluate and adjust the stability of the feedback mechanism based on the error trend model under fault conditions through regular simulation and actual flight tests.

[0064] In combination with all the above technical solutions, the beneficial effects of the present invention are as follows:

[0065] The present invention provides a rocket electrical system control method, which collects the flight status of the rocket and the external environment information; analyzes the flight status information through a data processing unit to identify potential abnormal situations; establishes a dynamic model of the rocket electrical system, and uses a dynamic model predictive control (MPC) algorithm to form an optimization control strategy for the parameters of the rocket electrical system according to the flight and environmental data updated in real time; establishes a feedback mechanism to adjust the gain matrix K in real time to ensure that the control strategy can adapt to the real-time changes of internal parameters and the external environment; and evaluates and adjusts the error handling strategy through regular simulation and actual flight tests. The present invention has high controllability and reliability, and can provide a more sophisticated and stable adjustment scheme for the control process of the rocket, thereby greatly improving the safety of the spacecraft and ground personnel.

[0066] The present invention combines the comprehensive application of multiple advanced technologies to significantly improve the flexibility and accuracy of the rocket electrical system in the actual control process. Through the application of the MPC algorithm, the present invention can accurately predict and adjust the control input in complex and changeable external conditions, effectively avoiding the out-of-control state caused by unknown factors. On the other hand, the timely automatic fault diagnosis function can immediately take over and respond to the fault when the system is disturbed and the flight is unstable, thereby ensuring the safety of the spacecraft. The present invention helps to save power resources to the greatest extent through the integration of intelligent energy management systems, thereby extending the service life of the spacecraft. At the same time, the automated control method avoids operational errors and safety risks that may be caused by human intervention. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] The accompanying drawings herein are incorporated in and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description, serve to explain the principles of the present disclosure;

[0068] Figure 1 is a flow chart of a rocket electrical system control method provided by an embodiment of the present invention;

[0069] Figure 2 is a schematic diagram of a rocket electrical system control device provided by an embodiment of the present invention;

[0070] In the figure: 1. Information collection module; 2. Potential abnormal situation identification module; 3. Dynamic model predictive control algorithm prediction module; 4. Feedback mechanism establishment module; 5. Evaluation and adjustment module. DETAILED DESCRIPTION

[0071] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without violating the connotation of the present invention, so the present invention is not limited by the specific implementation disclosed below.

[0072] Embodiment 1, as Figure 1 As shown, the rocket electrical system control method provided by the embodiment of the present invention includes:

[0073] S1, collects the rocket’s flight status and external environment information;

[0074] The flight status includes position, altitude, speed, acceleration, temperature, attitude and pressure, and the flight status is collected in real time by sensors and measuring equipment on the rocket and transmitted to the controller;

[0075] The external environment information includes factors that affect the aircraft and the route, such as weather conditions, terrain features, geological structures, etc.; exemplary examples include: air pressure, temperature, wind speed, atmospheric density, radiation level, and magnetic field strength;

[0076] The external environment information is acquired by sensors and environment detection equipment on the rocket and transmitted to the controller;

[0077] Examples include: physical properties such as wind direction, temperature, humidity, air pressure, altitude, atmospheric density, slope, etc. Meteorological conditions usually affect changes in air resistance, lift, inertia and thrust, thereby affecting important variables such as the location and speed of the landing point; taking into account the flight paths of surrounding spacecraft and their related navigation and control systems, and through the study of other spacecraft's flight paths, attitudes and other information, we can better understand the interaction between their orbits and avoid collision accidents.

[0078] S2, analyzes the flight status and external environment information through the data processing unit to identify potential abnormal situations;

[0079] Exemplarily, the flight status information collected by the sensor is preprocessed, and the future flight status is predicted based on the historical data and physical model of the rocket's flight status. The dynamic model predictive control algorithm is used to make short-term predictions of the position, speed, and attitude, etc. The flight status data is compared with the preset safety threshold. When a parameter exceeds the safety range, it is marked as a potential abnormality, such as the acceleration exceeds the set value or the attitude angle has a large deviation;

[0080] Exemplarily, it is also determined whether the external environmental information affects the normal flight of the rocket, mainly including whether the weather information, terrain information and geological structure information affect the normal flight of the rocket. The focus of weather information is to measure the accuracy and completeness of meteorological data in order to determine whether it is suitable for rocket takeoff and landing operations. Terrain information is mainly used to study the terrain distribution pattern, gradient fluctuation and other related properties, such as rock hardness, etc. The last step is to study the nature of the geological structure, especially the part involving rock and soil mechanical properties, which helps to confirm the land quality level and soil type, and determine its impact on the rocket. If there is no abnormal phenomenon, it goes to the next step S3; otherwise, it goes to step S4.

[0081] S3, based on the identification of potential abnormal conditions, a dynamic model of the rocket electrical system is established, and the dynamic model predictive control MPC algorithm is used to form an optimization control strategy for the rocket electrical system parameters according to the real-time updated flight status and environmental data;

[0082] The dynamic model predictive control algorithm uses the following formula to calculate the control input, which is input to the rocket power management module, thermal management module, attitude control module and communication and navigation module:

[0083] μ(t)=K×(x ref -x(t))+Δμ

[0084] Where μ(t) is the control input at time t, K is the gain matrix, and x ref is the reference trajectory state, x(t) is the current state, Δμ is the increment obtained by error adjustment, and Δμ is used to cope with real-time changes in the external environment and internal state;

[0085] Exemplarily, the errors are state error, prediction error and control input smoothing error, the state error is the difference between the current state of the system and the target state, the prediction error is the difference between the expected system state and the reference trajectory in the future prediction window of the dynamic model predictive control MPC algorithm, and the control input smoothing error is the change between the control input of the previous time step and the increment of the current time step to cope with the real-time changes of the external environment and internal state;

[0086] Among them, the rocket power management module, thermal management module, attitude control module and communication and navigation module use unified input, and take the weighted average of the state error, prediction error and control input smooth error at time t, and subtract the weighted average of the state error, prediction error and control input smooth error at the previous moment, that is, time t-1, to obtain the error increment, which is the increment obtained according to the error adjustment.

[0087] In another exemplary embodiment, the dynamic model predictive control algorithm can obtain the optimal solution through the least square method; the specific steps are as follows:

[0088] The first step is to obtain the state signal at each time point in the future prediction time window;

[0089] The second step is to construct the control input formula using linear regression or nonlinear mapping methods, that is, the state-error relationship function at all future time points;

[0090] The third step is to estimate the correlation coefficient and obtain the minimum value of the cost function;

[0091] The fourth step is to find the optimal solution, that is, the optimal control amount, based on the minimum value;

[0092] Step 5: Update the current state according to the optimal solution and determine the new control input;

[0093] Among them, the core of the dynamic model predictive control algorithm is to first construct a series of mathematical models of the relationship between flight status and electromagnetic signal xt (current state variable) response from past historical observation data, and then use the estimated value of this model as the target output, and adopt an appropriate cost function (optimization algorithm) based on minimizing the prediction error to solve the desired minimization control problem, and finally obtain a control signal output that meets the requirements.

[0094] Exemplarily, the dynamic model predictive control algorithm may also include state prediction for each time point within the future prediction time window, and use the cost function of minimizing the prediction error as the optimization criterion. The minimization deviation term can be achieved by introducing a self-correcting factor of the disturbance and continuously improved through an iterative update process. In addition, if large-scale shocks or disturbances that are difficult to eliminate are encountered, you can consider adding auxiliary controllers and connecting them to the main controller to achieve more accurate target positioning functions, which can effectively reduce deviations caused by external interference factors, thereby improving the performance of the entire control system relative to traditional methods.

[0095] In another exemplary embodiment, in step S3, the dynamic model predictive control algorithm includes performing state prediction for each time point in the future prediction time window, and adjusting the control input based on a cost function that minimizes the prediction error, wherein the cost function includes a weighted sum of the state error and the control input change, specifically:

[0096]

[0097] Where J is the weighted sum of the cost function, x(i|t) is the state from time i to time t, T is the order, μ(i|t) is the control input from time i to time t, μ(i-1|t) is the control input from the time before time i to time t, N is the length of the prediction window, and Q and R are weighted matrices used to balance prediction accuracy and control smoothness.

[0098] The control input μ(t) is obtained by minimizing the value of J, so that it fits the reference trajectory and reduces excessive adjustments while ensuring system stability. The trade-off between state error and control smoothness is adjusted through weighted matrices Q and R. The combination of state error and control input changes in the cost function enables MPC to have anti-interference capabilities in the following aspects: by reasonably selecting Q and R, the system can respond smoothly to small disturbances and avoid excessive response that leads to system instability; deviations from the reference trajectory are discovered early, and the control input is fine-tuned at the current moment; for emergencies, the prediction mechanism of MPC plays a preventive role. When emergencies lead to error accumulation, the MPC system will adaptively adjust the gain matrix K, increase the feedback strength, and accelerate the adjustment in a short period of time to restore stability.

[0099] Exemplarily, in step S3, the optimization control strategy includes:

[0100] Under the premise of flight stability, the dynamic model predictive control algorithm predicts energy demand and adjusts electrical system parameters to minimize energy consumption;

[0101] The time response to the external disturbance is smoothed by selecting the weighting matrices Q and R of the cost function, so that the response time is adapted to the external disturbance;

[0102] Setting error ratios or safety constraints for a certain period of time to prevent electrical system failures and ensure flight safety; in preventing electrical system failures, real-time analysis of electrical system performance data is performed, and possible system failures are identified based on a predetermined dynamic model predictive control algorithm, and the system automatically switches to an emergency backup plan or notifies the ground control center; when an abnormal situation occurs, an alarm signal is sent in a timely manner, and the remote control terminal intervenes and takes emergency measures for the corresponding components. For example, some early warning thresholds are set to remind operators to make preventive preparations in advance. For example, when the error exceeds a certain ratio or lasts for a period of time, the operator needs to stop working immediately and investigate the cause to avoid irreversible consequences.

[0103] Exemplarily, minimizing energy consumption includes optimizing the management of the rocket's energy consumption, predicting energy demand through a dynamic model predictive control algorithm, and adjusting electrical system parameters to maximize energy efficiency.

[0104] Preventing electrical system failures includes automatic fault diagnosis, which analyzes the performance data of the electrical system in real time and identifies possible system failures based on a predetermined dynamic model predictive control algorithm, and automatically switches to an emergency backup plan or notifies the ground control center when necessary.

[0105] When an abnormal situation occurs, an alarm signal is sent in time, and the remote control end intervenes and takes emergency measures for the corresponding components, such as replacing parts or temporarily taking over control authority, etc., to reduce the risk of loss. In another optional example, the above steps can be repeated at appropriate time intervals based on actual conditions, and with each improvement, the gap between the initial time and the subsequent time is gradually reduced until the best effect is achieved.

[0106] S4, establish a feedback mechanism based on the error trend model to adjust the gain matrix K in real time to ensure that the optimization control strategy can adapt to the real-time changes of internal parameters and external environment;

[0107] Exemplarily, the feedback mechanism based on the error trend model collects data in real time and calculates the error, analyzes the error, and identifies its trend. If the error increases significantly, the system deviation is large, and it is necessary to increase the relevant value of K to increase the control input. The updated value of the gain matrix K is calculated using the aforementioned cost function (optimization algorithm) based on minimizing the prediction error. The updated gain matrix K is used to calculate the new control input μ(t), and is sent to the relevant controller of the rocket electrical system.

[0108] Building an error trend model includes:

[0109] S201, quantitative calculation of the incremental values ​​of the single factors of flight status data, environmental data, error change data and fault information obtained according to the error adjustment, comprehensively considering the values ​​of the three dimensions of simulated state error weight, prediction error weight and control input smoothing error, and assigning a real value between 0 and 1 to each single factor simulation;

[0110] S202, error trend exponential smoothing linear line calculation of fault-related simulation, by calculating the deviation value PAI, the difference linear mean XYU and the high and low side distribution GDBF histogram of the incremental value obtained according to the error adjustment in different time periods, and calculating the error trend exponential smoothing moving linear line indicator HAD; HAD Mt and HAD Nt They are respectively the exponentially smoothed moving linear average of the error trend in the M time interval and the exponentially smoothed moving linear average of the error trend in the N time interval at the tth moment, where M>N;

[0111] S203, visualization of error trend prediction, after calculating the incremental value obtained by error adjustment and the PAI, XYU and PAIF values ​​in the specified time period and time cycle of the flight status data, environmental data, error change data and fault information single factor impact condition simulation, the visual window visualization framework is used to draw curves according to time periods with different weights, and the intersection points of the fast and slow lines and the points crossing the horizontal axis are highlighted, and the overall change trend of multiple curves and the distribution of key points are provided to assist users to quickly grasp the development and evolution of the entire error and the potential change trend in the future;

[0112] The step S201 specifically includes:

[0113] (1) Setting of state error weight: The state error weight is divided into 10 levels according to the fault severity ranking. The final score of the state error weight is recorded as G1;

[0114] (2) state error prediction error weight. The type of state error prediction error is automatically determined using the state tag. The state error prediction error weight is recorded as G2.

[0115] (3) Control input smoothing error weight: reasonably assign a smoothing coefficient to the control input smoothing error according to the simulation.

[0116] The method (3) specifically includes: automatically extracting the simulation links in the sliding window where the simulation is located; automatically sorting the links of the simulation type according to their appearance positions in the error source code; letting T represent the number of all error links, lianf c represents the links sorted as c, then the fth c Importance calculation of links:

[0117]

[0118] In the formula, G(lianf c ) is the f c The important value of a link, c is the sequence number;

[0119] Change the link importance to a value between 50 and 100 as the control input smoothing error weight, denoted as G3, and the expression is:

[0120] G3=sigmoid(G(lianf c ))×100.

[0121] Step S202 specifically includes:

[0122] (1) Calculate the HAD value using an exponentially decreasing weighted moving average. The expression is:

[0123]

[0124] In the formula, HAD Nt-1 is the exponentially smoothed moving linear average of the error trend in N time intervals at time t-1, N is the length of the prediction window, and A t is the exponential smoothing moving linear constant at the tth moment, M is the Mth time interval, HAD Mt-1 (M-1) is the exponentially smoothed moving linear average of the error trend in the M time interval at time t-1;

[0125] (2) Calculate the PAI value, the HAD value of the N time interval minus the HAD value of the M time interval. When the error spreads more, the HAD of the N time interval is higher than the HAD of the M time interval. The expression is:

[0126] PAI t =HAD Nt -HAD Mt

[0127] Where PAI t is the PAI value at time t;

[0128] (3) Calculate the XYU value, which is the time interval error trend exponential smoothing moving linear line of the PAI line. The expression is:

[0129]

[0130] In the formula, XYU pt is the XYU value at time t, XYU pt-1 is the XYU value at time t-1, and p is a time different from time t;

[0131] (4) Calculate the moving smoothing trend value, and let the moving smoothing trend value represent the overall development trend of the error. The current trend value is calculated using the high and low edge distribution, and the expression is:

[0132] GDBF t =2(PAI t -XYU pt )

[0133] In the formula, GDBF t is the moving smoothing trend value at time t;

[0134] A crossover signal occurs when the shorter-period HAD crosses the longer-period HAD.

[0135] As another example, in step S4, the feedback mechanism monitors the execution effect of the optimized control signal in real time, compares the monitoring result with the reference trajectory, calculates the error, and adjusts the control input according to the error to improve the control accuracy and ensure the stability of the system.

[0136] Among them, the feedback mechanism is a cyclic process, that is, continuously acquiring real-time flight status and external environment data, issuing instructions to hardware devices such as motors / electronic control units or chip processors after data analysis, and then converting them into specific actions through actuators, thereby achieving the effect of controlling the parameters of the rocket's electrical system. In this process, it is also necessary to continuously correct and adjust the control input to make the regulation process as close to the expected ideal level as possible.

[0137] S5, evaluate and adjust the stability of the feedback mechanism based on the error trend model under fault conditions through regular simulation and actual flight tests.

[0138] Exemplarily, the simulated flight test obtains flight status data, control response data, environmental change data and system error data, and the actual flight test obtains actual flight status data, environmental data, error change data and fault information;

[0139] The data obtained from the simulated flight test is compared with the preset reference trajectory to analyze the response effects of different optimization control strategies. The error trend model is established using the error data to analyze the error change law. The gain, feedback frequency, response speed and other parameters of the optimization control strategy are adjusted according to the error trend model, and the influence of each parameter on the error convergence speed is tested. The simulation test results are used to adjust and optimize the control parameters in the feedback mechanism based on the error trend model, such as the initial value of the gain matrix K. The data obtained from the actual flight test uses the error change trend over time to judge the adaptability of the existing feedback mechanism (error processing strategy) based on the error trend model. The simulation data is compared with the actual flight data to analyze whether the parameters in the simulation need to be adjusted. The reliability of the feedback mechanism based on the error trend model under fault conditions is evaluated through the analysis of fault data. According to the error feedback results of the actual flight, the trigger threshold and response time of the emergency optimization control strategy are optimized.

[0140] Embodiment 2, as Figure 2 As shown, an embodiment of the present invention provides a rocket electrical system control device, comprising:

[0141] Information collection module 1, used to collect the flight status of the rocket and external environment information;

[0142] Potential abnormal situation identification module 2, used for analyzing the flight status and external environment information through the data processing unit to identify potential abnormal situations;

[0143] Dynamic model predictive control algorithm prediction module 3 is used to establish a dynamic model of the rocket electrical system based on the identification of potential abnormal conditions, and use the dynamic model predictive control MPC algorithm to form an optimization control strategy for the parameters of the rocket electrical system according to the real-time updated flight status and environmental data;

[0144] Feedback mechanism establishment module 4 is used to establish a feedback mechanism based on the error trend model, adjust the gain matrix K in real time, and make the optimization control strategy adapt to the real-time changes of internal parameters and external environment;

[0145] The evaluation and adjustment module 5 is used to evaluate and adjust the stability of the feedback mechanism based on the error trend model under fault conditions through regular simulation and actual flight tests.

[0146] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0147] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with the technical field within the technical scope disclosed by the present invention and within the spirit and principles of the present invention should be covered within the protection scope of the present invention.

Claims

1. A rocket electrical system control method, characterized in that: The method comprises the following steps: S1, collects the rocket’s flight status and external environment information; S2, analyzes the flight status and external environment information through the data processing unit to identify potential abnormal situations; S3, based on the identification of potential abnormal conditions, a dynamic model of the rocket electrical system is established, and the dynamic model predictive control MPC algorithm is used to form an optimization control strategy for the rocket electrical system parameters according to the real-time updated flight status and environmental data; S4, establish a feedback mechanism based on the error trend model to adjust the gain matrix K in real time so that the optimization control strategy can adapt to the real-time changes of internal parameters and external environment; S5, evaluate and adjust the stability of the feedback mechanism based on the error trend model under fault conditions through regular simulation and actual flight tests.

2. The rocket electrical system control method according to claim 1, characterized in that: In step S1, the flight status is collected in real time by sensors and measuring equipment on the rocket and transmitted to the controller. The flight status includes: position, altitude, speed, acceleration, temperature, attitude and pressure; The external environmental information is acquired by sensors and environmental detection equipment on the rocket and transmitted to the controller. The external environmental information includes factors such as meteorological conditions, terrain features, and geological structures that affect the aircraft and routes; specifically including: air pressure, temperature, wind speed, atmospheric density, radiation levels, and magnetic field strength.

3. The rocket electrical system control method according to claim 1, characterized in that: In step S2, the data processing unit pre-processes the flight status information collected by the sensor, predicts the future flight status based on the historical data and physical model of the rocket's flight status, uses the dynamic model predictive control algorithm to make short-term predictions on the position, speed and attitude, and compares the flight status data with the preset safety threshold. When a parameter exceeds the safety range, it is marked as a potential abnormality; The dynamic model predictive control algorithm calculates the control input using the following formula: μ(t)=K×(x ref -x(t))+Dm Where μ(t) is the control input at time t, K is the gain matrix, and x ref is the reference trajectory state, x(t) is the current state, Δμ is the increment obtained by error adjustment, and Δμ is used to cope with real-time changes in the external environment and internal state; The control input is input to the rocket power management module, thermal management module, attitude control module and communication and navigation module; the errors are state error, prediction error and control input smoothing error, the state error is the difference between the current state of the system and the target state, the prediction error is the difference between the expected system state and the reference trajectory in the future prediction window of the dynamic model predictive control MPC algorithm, and the control input smoothing error is the change between the control input of the previous time step and the increment of the current time step, which responds to real-time changes in the external environment and internal state; The rocket power management module, thermal management module, attitude control module and communication and navigation module use unified input, take the weighted average of the state error, prediction error and control input smooth error at time t, and subtract the weighted average of the state error, prediction error and control input smooth error at the previous time, i.e., time t-1, to obtain the error increment, i.e., the increment obtained according to the error adjustment.

4. The rocket electrical system control method according to claim 3, characterized in that: The dynamic model predictive control algorithm obtains the optimal solution through the least squares method. The following steps are involved: The first step is to obtain the state signal at each time point in the future prediction time window; The second step is to construct the control input formula using linear regression or nonlinear mapping methods to obtain the state-error relationship function at all future time points; The third step is to estimate the state error and the control input variation coefficient to obtain the minimum value of the cost function; The fourth step is to find the optimal solution based on the minimum value and obtain the optimal control amount; The fifth step is to update the current state according to the optimal solution and determine the new control input.

5. The rocket electrical system control method according to claim 4, characterized in that: In the third step, the cost function consists of a weighted sum of the state error and the control input change, expressed as: Where J is the weighted sum of the cost function, x(i|t) is the state from time i to time t, T is the order, μ(i|t) is the control input from time i to time t, μ(i-1|t) is the control input from the time before time i to time t, N is the length of the prediction window, and Q and R are weighted matrices used to balance prediction accuracy and control smoothness.

6. The rocket electrical system control method according to claim 1, characterized in that: In step S3, the optimization control strategy includes: Under the premise of flight stability, the dynamic model predictive control algorithm predicts energy demand and adjusts electrical system parameters to minimize energy consumption; The time response to the external disturbance is smoothed by selecting the weighting matrices Q and R of the cost function, so that the response time is adapted to the external disturbance; Setting an error ratio or a continuous time interval safety constraint can prevent electrical system failures and ensure flight safety; in preventing electrical system failures, the performance data of the electrical system is analyzed in real time, and possible system failures are identified based on a predetermined dynamic model predictive control algorithm, and the system automatically switches to an emergency backup plan or notifies the ground control center; when an abnormal situation occurs, an alarm signal is sent in a timely manner, and the remote control terminal intervenes and takes emergency measures for the corresponding components.

7. The rocket electrical system control method according to claim 1, characterized in that: In step S4, the feedback mechanism based on the error trend model is established to adjust the gain matrix K in real time, including: The feedback mechanism based on the error trend model collects data in real time and calculates the error, analyzes the error, and identifies the trend. If the error increases significantly and the system deviation is large, the relevant value of K is increased to increase the control input, and the updated value of the gain matrix K is calculated using the aforementioned cost function. The updated gain matrix K is used to calculate the new control input μ(t) and send it to the controller of the rocket electrical system.

8. The rocket electrical system control method according to claim 1, characterized in that: In step S5, the stability of the feedback mechanism based on the error trend model under fault conditions is evaluated and adjusted, including: regularly simulating flight tests to obtain flight status data, control response data, environmental change data and system error data, and actually performing flight tests to obtain actual flight status data, environmental data, error change data and fault information; comparing the data obtained from the simulated flight test with the preset reference trajectory, analyzing the response effects of different optimization control strategies, establishing an error trend model using the error data, analyzing the error change law, adjusting the gain, feedback frequency and response speed parameters of the optimization control strategy according to the error trend model, and testing the influence of each parameter on the error convergence speed, and the simulation test results are used to adjust and optimize the control parameters such as the initial value of the gain matrix K in the feedback mechanism based on the error trend model; using the data obtained from the actual flight test to determine the adaptability of the existing feedback mechanism based on the error trend model using the error change trend over time, comparing the simulated data with the actual flight data, analyzing whether the parameters in the simulation need to be adjusted, evaluating the reliability of the feedback mechanism based on the error trend model under fault conditions through the analysis of the fault data, and optimizing the trigger threshold and response time of the emergency optimization control strategy according to the error feedback results of the actual flight.

9. The rocket electrical system control method according to claim 8, characterized in that: Use error data to build an error trend model, including: Step 1, quantitative calculation of the incremental value obtained by error adjustment for single factors of flight status data, environmental data, error change data and fault information, comprehensive consideration of the values ​​of three dimensions of simulated state error weight, prediction error weight and control input smooth error, assign a real value between 0 and 1 to each single factor simulation; setting of state error weight, dividing the state error weight into 10 steps according to the fault severity ranking, the final score of the state error weight is recorded as G1; state error prediction error weight, the type of state error prediction error is automatically determined by state marking, the state error prediction error weight is recorded as G2; control input smooth error weight, reasonably assigning its smoothing coefficient according to the simulated control input smooth error; Step 2, the error trend exponential smoothing linear line calculation of the fault-related simulation, by calculating the deviation value PAI, the difference linear mean XYU and the high and low side distribution GDBF histogram of the incremental value obtained according to the error adjustment in different time periods, and calculating the error trend exponential smoothing moving linear line indicator HAD; HAD Mt and HAD Nt They are respectively the exponentially smoothed moving linear average of the error trend in the M time interval and the exponentially smoothed moving linear average of the error trend in the N time interval at the tth moment, where M>N; (1) Calculate the HAD value using an exponentially decreasing weighted moving average. The expression is: In the formula, HAD Nt-1 is the exponential smoothing moving linear average of the error trend in N time intervals at time t-1, N is the length of the prediction window, and A t is the exponential smoothing moving linear constant at the tth moment, M is the Mth time interval, HAD Mt-1 (M-1) is the exponentially smoothed moving linear average of the error trend in the M time interval at time t-1; (2) Calculate the PAI value, the HAD value of the N time interval minus the HAD value of the M time interval. When the error spreads more, the HAD of the N time interval is higher than the HAD of the M time interval. The expression is: PAI t =HAD Nt --HAD Mt Where PAI t is the PAI value at time t; (3) Calculate the XYU value, which is the time interval error trend exponential smoothing moving linear line of the PAI line. The expression is: In the formula, XYU pt is the XYU value at time t, XYU pt-1 is the XYU value at time t-1, and p is a time different from time t; (4) Calculate the moving smoothing trend value, and let the moving smoothing trend value represent the overall development trend of the error. The current trend value is calculated using the high and low edge distribution, and the expression is: GDBF t =2(PAI t -XYU pt ) In the formula, GDBF t is the moving smoothing trend value at time t; A crossover signal occurs when the shorter-period HAD crosses the longer-period HAD. Step 3, visualization of error trend prediction, after calculating the incremental value obtained by error adjustment and the PAI, XYU and PAIF values ​​in the specified time period and time cycle of the single factor impact condition simulation of flight status data, environmental data, error change data and fault information, use the visual window visualization framework to draw curves according to time periods with different weights, and highlight the intersection points of fast and slow lines and the points crossing the horizontal axis. By providing the overall change trend of multiple curves and the distribution of key points, it helps users quickly grasp the development and evolution of the entire error and the potential change trend in the future; By automatically extracting the simulation links of the sliding window where the simulation is located; automatically sorting the links of the simulation type in the error source code; let T represent the number of all error links, lianf c represents the links sorted as c, then the fth c Importance calculation of links: In the formula, G(lianf c ) is the f c The important value of a link, c is the sequence number; Change the link importance to a value between 50 and 100 as the control input smoothing error weight, denoted as G3, and the expression is: G3=sigmoid(G(linaf c ))×100。 10. A rocket electrical system control device, characterized in that: The device implements a rocket electrical system control method as claimed in any one of claims 1 to 9, and the device comprises: An information collection module (1) is used to collect the flight status of the rocket and external environment information; A potential abnormal situation identification module (2), used for analyzing the flight status and external environment information through a data processing unit to identify potential abnormal situations; A dynamic model predictive control algorithm prediction module (3) is used to establish a dynamic model of the rocket electrical system based on the identification of potential abnormal conditions, and to form an optimization control strategy for the parameters of the rocket electrical system according to the flight status and environmental data updated in real time using the dynamic model predictive control MPC algorithm; A feedback mechanism establishment module (4) is used to establish a feedback mechanism based on an error trend model, adjust the gain matrix K in real time, and make the optimization control strategy adapt to the real-time changes of internal parameters and external environment; An evaluation and adjustment module (5) is used to evaluate and adjust the stability of the feedback mechanism based on the error trend model under fault conditions through regular simulation and actual flight tests.

Images