Unmanned aerial vehicle flight control system and method based on multi-source sensor fusion and adaptive decision

Abstract

The invention discloses an unmanned aerial vehicle flight control system and method based on multi-source sensor fusion and adaptive decision, relates to the field of unmanned aerial vehicle control, and solves the problem of insufficient stability of existing unmanned aerial vehicle flight control. The data processing module is used for processing the environment parameter matrix to obtain a dynamic model of the unmanned aerial vehicle; performing parameter prediction on the unmanned aerial vehicle to obtain a prediction value, and performing linear recursion on the unmanned aerial vehicle parameter matrix and the prediction value to obtain an optimal parameter matrix of the unmanned aerial vehicle; the monitoring control module is used for performing deviation calculation on the real-time data of the unmanned aerial vehicle and the optimal parameter matrix, performing control judgment on the unmanned aerial vehicle and determining the running state of the unmanned aerial vehicle; the self-adaptive decision module is used for carrying out self-adaptive adjustment on the real-time data of the unmanned aerial vehicle and storing and recording a self-adaptive adjustment result; the stability of flight control of the unmanned aerial vehicle can be effectively enhanced, and the applicability of the unmanned aerial vehicle is improved.

Description

Technical Field

[0001] This invention belongs to the field of unmanned aerial vehicle (UAV) flight control and relates to UAV technology, specifically a UAV flight control system and method based on multi-source sensor fusion and adaptive decision-making. Background Technology

[0002] Existing UAV flight control systems and methods have the following specific shortcomings when performing flight control:

[0003] 1. Most existing UAV flight control systems rely on a single sensor for control. When using a single positioning system (such as the Global Positioning System or the BeiDou Navigation Satellite System), the UAV is susceptible to interference in complex geological environments, leading to inaccurate positioning. Furthermore, when the sensor malfunctions, the UAV is significantly affected, making it impossible to effectively control the UAV.

[0004] 2. Traditional drone control methods are often difficult to adapt to complex and ever-changing environments. In dynamic environments, situations such as changes in wind force, the sudden appearance of other drones, birds, or other obstacles can greatly affect the flight stability of drones, resulting in low applicability.

[0005] To this end, we propose an unmanned aerial vehicle (UAV) flight control system and method based on multi-source sensor fusion and adaptive decision-making. Summary of the Invention

[0006] To address the shortcomings of existing technologies, the present invention aims to provide a UAV flight control system and method based on multi-source sensor fusion and adaptive decision-making. The present invention aims to enhance the stability of UAV flight control and improve the applicability of UAVs.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: a UAV flight control system based on multi-source sensor fusion and adaptive decision-making, the specific working process of each module is as follows:

[0008] Data acquisition module: Acquires flight mission information, uses the UAV based on the flight mission information, acquires the UAV's operating parameters and external environment, and obtains UAV information;

[0009] Data processing module: Based on UAV information, obtain the environmental parameter matrix, process the environmental parameter matrix to obtain the UAV dynamic model; use the UAV dynamic model to predict UAV parameters and obtain predicted values; obtain the UAV parameter matrix at different times through the environmental parameter matrix; perform optimal linear recursive estimation between the UAV parameter matrix at different times and the predicted values ​​to obtain the optimal parameter matrix of the UAV.

[0010] Monitoring and control module: Acquires real-time data of the UAV, calculates the deviation between the real-time data and the optimal parameter matrix to obtain the deviation value of the UAV, and makes control judgments on the UAV based on the deviation value to determine the operating status of the UAV.

[0011] Adaptive Decision Module: The adaptive decision module determines the operating status of the UAV, adaptively adjusts the real-time data of the UAV, judges the adaptive adjustment results by combining the optimal parameter matrix, and stores and records the adaptive adjustment results.

[0012] Furthermore, the operating parameters of the drone and the external environment are acquired, as follows:

[0013] Obtain the flight origin, construct a three-dimensional coordinate system from the flight origin, and represent points on the three-dimensional coordinate system through the x, y, and z axes; measure the UAV parameters during UAV movement, specifically including:

[0014] The flight deflection angle and angular velocity of the UAV, the acceleration of the UAV in the x, y, and z axes, the magnetic force of the UAV on the UAV in the x, y, and z axes, and the position parameters of the UAV are measured.

[0015] The UAV parameters are statistically analyzed to obtain the UAV parameter matrix cs, cs = [csy1, csy2, ..., csy...]. a ]; where 'a' represents the number of UAV parameters, and 'csy' represents the number of parameters. a This represents the parameters of the a-th UAV;

[0016] Based on the temporal sequence of UAV parameter measurements, the UAV parameter matrices at different times are obtained and denoted as cs(1) to cs(b).

[0017] The external environment in which the drone is located is detected, and the type of external environment in which the drone is located is obtained, denoted as c. Based on the type of external environment in which the drone is located, the parameter matrix of the drone is integrated to obtain the environmental parameter matrix hcj.

[0018] Furthermore, the environmental parameter matrix is ​​processed as follows:

[0019] Based on the environmental parameter matrix, the parameter matrices of UAVs under the same environment are obtained to obtain the UAV's time-series parameter matrix; based on the UAV's time-series parameter matrix, the UAV parameter matrix at different times is obtained as cs(i,j)=[csy1(i,j), csy2(i,j), ..., csy a [(i, j)];

[0020] The difference between the UAV parameters at adjacent different times is calculated to obtain the parameter change matrix cbh;

[0021] Based on the parameter change matrix, obtain the parameter change value for each parameter and denote it as cbh1(i,j) to cbh a (i, j);

[0022] The parameter changes at different times are statistically analyzed, and the number of parameter changes is denoted as zs. The number of parameter changes that are not equal to 0, bs, is obtained. Combined with the number of parameter changes, zs, the parameter change frequency is calculated to obtain the parameter change frequency bhp(j).

[0023] Based on the calculation process of the parameter change frequency, the parameter change frequency of each parameter is denoted as bhp1(j) to bhp. a (j); bhp a (j) represents the frequency of parameter change of the a-th parameter under the j-th environment;

[0024] By combining the parameter change value with the parameter change frequency, the stable value of the parameter is calculated to obtain the stable value wdz; the parameter is judged based on the stable value.

[0025] Furthermore, the parameters are judged based on their stable values, as follows:

[0026] The mean of parameter changes is calculated, and the mean of parameter changes is used as the threshold yz.

[0027] The stable value of parameter wdz is determined by the threshold yz;

[0028] When wdz ≤ yz, it indicates that the parameters are relatively stable, and the parameters should be saved.

[0029] When wdz > yz, it indicates that the parameters are relatively unstable, and the parameters need to be processed.

[0030] Furthermore, the parameters are processed as follows:

[0031] For parameters that are unstable, we obtain them, denoted as bwd1(j) to bwd. ab (j); Based on the temporal sequence of the parameters, the parameters at all times within time b are obtained, denoted as bwd1(i,j) to bwd. ab (i, j);

[0032] Transfer bwd1(i,j) to bwd ab Processing (i, j) yields its regression function, denoted as FH1(i) to FH1(i). ab (i); where FH ab(i) represents the regression value of the ab-th parameter at time i;

[0033] Parameters that are relatively stable are represented by constants;

[0034] Integrate all parameters, through F1(i,j) to F a Representing (i, j) yields the dynamic model dmx of the UAV;

[0035] Based on the dynamic model of the UAV, dynamic prediction is performed on the UAV to obtain the predicted value dmx(i,j). The predicted value dmx(i,j) is combined with the environmental parameter matrix to perform optimal linear recursive estimation of the UAV's flight state, and the optimal parameter matrix of the UAV's flight state is obtained.

[0036] Furthermore, the deviation between the UAV's real-time data and the optimal parameter matrix is ​​calculated, as follows:

[0037] Obtain the optimal parameter matrix zuy for the UAV's flight state. Let us denote the number of parameters in the optimal parameter matrix as us, and the parameters in the optimal parameter matrix as zss; zuy = [zss1, zss2, ..., zss us ];

[0038] The real-time parameters of the UAV are acquired and denoted as ssc; the real-time parameters of the UAV are statistically analyzed to obtain the real-time parameter matrix scj; the data format of the real-time parameter matrix of the UAV is made consistent with that of the optimal parameter matrix to obtain the real-time parameter matrix scj = [ssc1, ssc2, ..., ssc us ];

[0039] The correlation values ​​of the matrix parameters in the real-time parameter matrix of the UAV are determined, and the parameters in the real-time parameter matrix are iteratively checked from ssc1 to ssc. us The parameters are adjusted sequentially, and the correlation values ​​glz1 to glz of the matrix parameters are obtained based on the number of other parameters that change simultaneously when a parameter is adjusted. us ;

[0040] Based on the correlation values, the associated matrix parameters in the real-time parameter matrix are obtained to obtain the correlation parameters; denoted as xgl1 to xgl xg ; Combining the optimal matrix parameters, for xgl1 to xgl xg Obtain the corresponding parameters to get the optimal corresponding parameters; denoted as xzy1 to xzy xg ;

[0041] Based on the correlation parameters and the optimal corresponding parameters, the influence value of each parameter is calculated to obtain the influence value yxz. u ;

[0042]

[0043] Among them: glz u This represents the associated value of the u-th matrix parameter; xgl v xgl represents the v-th associated parameter. v For the vth Optimal corresponding parameters;

[0044] The deviation value is calculated by using the influence value as a weight and combining the real-time data of the UAV with the optimal parameter matrix.

[0045] Furthermore, the deviation value is calculated as follows:

[0046] Based on the real-time parameter matrix of the UAV [ssc1, ssc2, ..., ssc us ] and the optimal parameter matrix [zss1, zss2, ..., zss us The parameters in the matrix are calculated, and the influence values ​​y, x, and z of the matrix parameters are combined. u The deviation value of the drone is calculated to obtain the deviation value plz;

[0047]

[0048] Among them: ssc u zss represents the u-th parameter in the real-time parameter matrix. u This represents the u-th parameter in the optimal parameter matrix;

[0049] The deviation values ​​of the previous h times are calculated according to the formula of the deviation value plz, and plz(1) to plz(h) are obtained; the state of the UAV is judged by the deviation values ​​plz(1) to plz(h).

[0050] Furthermore, the drone's status is assessed, specifically as follows:

[0051] The mean deviation value plj is obtained by calculating the mean deviation value of the deviation values ​​of the first h time points;

[0052] The mean change of the deviation value over the first h time steps is calculated to obtain the mean change value bhj;

[0053] Based on the deviation from the mean plj and the change mean bhj, the current deviation value is used to determine the state and obtain the current state value ztz.

[0054]

[0055] Where: plz(h) represents the deviation value of the previous time step at the current time step;

[0056] If the state value ztz is less than 0, it indicates that the parameters of the UAV are far from the optimal parameter matrix and manual intervention is required.

[0057] If the state value ztz is greater than or equal to 0, it indicates that the parameters of the UAV are consistent with the optimal parameter matrix, or that the parameters of the UAV are changing toward the optimal parameter matrix; this is suitable for UAVs to make adaptive decisions.

[0058] Furthermore, adaptive adjustments are made to the real-time data of the drone, as follows:

[0059] Once it is determined that the UAV can make adaptive decisions, based on the UAV's real-time data and the optimal parameter matrix, the UAV makes adaptive decision judgments to fit the UAV's adaptive decision results with the optimal parameter matrix, and the various parameters of the UAV's operating state tend to the optimal parameter matrix; different UAV adaptive adjustment results are saved, and the saved data is transmitted during the next UAV adaptive adjustment, and the UAV makes adaptive decisions based on the existing adaptive adjustment data.

[0060] A UAV flight control method based on multi-source sensor fusion and adaptive decision-making, the control method includes:

[0061] Step S1: Data acquisition module: acquires flight mission information, uses the UAV based on the flight mission information, acquires the UAV's operating parameters and external environment, and obtains UAV information;

[0062] Step S2: Based on the UAV information, obtain the environmental parameter matrix, process the environmental parameter matrix to obtain the UAV dynamic model; use the UAV dynamic model to predict the parameters of the UAV and obtain the predicted values; obtain the UAV parameter matrix at different times through the environmental parameter matrix; perform optimal linear recursive estimation between the UAV parameter matrix at different times and the predicted values ​​to obtain the optimal parameter matrix of the UAV.

[0063] Step S3: Acquire real-time data of the UAV, calculate the deviation between the real-time data of the UAV and the optimal parameter matrix to obtain the deviation value of the UAV, and make control judgments on the UAV based on the deviation value to determine the operating status of the UAV.

[0064] Step S4: Determine the operating status of the UAV as an adaptive decision module, adaptively adjust the real-time data of the UAV, judge the adaptive adjustment result by combining the optimal parameter matrix, and store and record the adaptive adjustment result.

[0065] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0066] 1. This invention integrates multi-source data from multiple sensors, giving full play to the advantages of each sensor and allowing them to cooperate to eliminate their respective defects, thereby improving the positioning accuracy and stability of the UAV in complex environments. If one or more sensors in the sensing system are interfered with or fail to work properly, other sensors can continue to provide data to the flight control system, ensuring the normal operation of the flight control system.

[0067] 2. By employing a Kalman filter and integrating comprehensive sensor data, the system predicts the optimal flight state for the current flight environment and maintains an accurate estimate of the UAV's state, thereby ensuring flight stability. Simultaneously, the system makes adaptive decisions for the UAV based on the prediction results of the Kalman filter, enhancing the UAV's adaptability to complex and changing environments. By analyzing the UAV's state and environmental factors in real time and adaptively adjusting flight parameters, the UAV can more flexibly cope with challenges such as wind changes and obstacle avoidance, ensuring flight safety and mission execution efficiency.

[0068] 3. During data calculation, parameters are calculated from multiple dimensions to reduce calculation errors. At the same time, the flight deviation and flight status of the drone are calculated in detail and displayed intuitively, so that users can discover and deal with flight problems in a timely manner, thereby enhancing the flight safety of the drone. Attached Figure Description

[0069] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0070] Figure 1 This is an overall system block diagram of the present invention;

[0071] Figure 2 This is a schematic diagram of the data processing of the present invention;

[0072] Figure 3 This is a schematic diagram of the method of the present invention; Detailed Implementation

[0073] 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 of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0074] Example 1

[0075] Please see Figure 1The present invention provides a technical solution: a UAV flight control system based on multi-source sensor fusion and adaptive decision-making, comprising a data acquisition module, a data processing module, a monitoring and control module, an adaptive decision-making module and a server, wherein the data acquisition module, the data processing module, the monitoring and control module and the adaptive decision-making module are respectively connected to the server, and the server controls the data acquisition module, the data processing module, the monitoring and control module and the adaptive decision-making module respectively;

[0076] Data acquisition module: Acquires flight mission information, uses the UAV based on the flight mission information, acquires the UAV's operating parameters and external environment, and obtains UAV information;

[0077] The specific workflow of the data acquisition module is as follows:

[0078] Based on the flight mission information, the flight origin is obtained, and a three-dimensional coordinate system is constructed from the flight origin. Points on the three-dimensional coordinate system are represented by the x, y, and z axes. The flight path is planned according to the flight mission information, and the UAV moves according to the flight path. The UAV parameters are measured during the movement, as follows:

[0079] The drone's flight deflection angle and angular velocity are measured using gyroscopes and fiber optic gyroscopes; accelerometers are used to detect the drone's acceleration along the x, y, and z axes; and magnetometers are used to measure the magnetic force exerted on the drone by the Earth's magnetic field along the x, y, and z axes. The drone's position parameters are acquired using GNSS combined with lidar; and the drone parameters are statistically analyzed to obtain the drone parameter matrix cs, cs = [csy1, csy2, ..., csy...]. a ]; where 'a' represents the number of UAV parameters, and 'csy' represents the number of parameters. a This represents the parameters of the a-th UAV;

[0080] Based on the temporal sequence of UAV parameter measurements, the UAV parameter matrices at different times are obtained and denoted as cs(1) to cs(b).

[0081] The external environment of the UAV is detected by a visible light camera and an infrared thermal imager, and the type of external environment of the UAV is obtained, denoted as c. Based on the type of external environment of the UAV, the parameter matrix of the UAV is integrated to obtain the environmental parameter matrix hcj.

[0082]

[0083] It should be noted that: cs(b, c) represents the UAV parameter matrix at time b when the UAV is in the c-th environment; each column of the environment parameter matrix represents the time-series parameter matrix of the UAV in the same environment.

[0084] Data processing module: Based on UAV information, obtain the environmental parameter matrix, process the environmental parameter matrix to obtain the UAV dynamic model; use the UAV dynamic model to predict UAV parameters and obtain predicted values; obtain the UAV parameter matrix at different times through the environmental parameter matrix; perform optimal linear recursive estimation between the UAV parameter matrix at different times and the predicted values ​​to obtain the optimal parameter matrix of the UAV.

[0085] The specific workflow of the data processing module is as follows:

[0086] Based on the environmental parameter matrix, the parameter matrices of UAVs under the same environment are obtained to obtain the UAV's time-series parameter matrix; based on the UAV's time-series parameter matrix, the UAV parameter matrix at different times is obtained as cs(i,j)=[csy1(i,j), csy2(i,j), ..., csy a [(i, j)];

[0087] It should be noted that: cs(i, j) represents the drone parameter matrix at time i when the drone is in the j-th environment; csy a (i, j) represents the a-th UAV parameter in the UAV parameter matrix at time i under the j-th environment.

[0088] The difference between the UAV parameters at adjacent different times is calculated to obtain the parameter change matrix cbh;

[0089] cbh(i,j)=cs(i,j)-cs(i-1,j);

[0090] It should be noted that by calculating the numerical changes of the drone, the flight status of the drone can be displayed intuitively, thus improving the accuracy of drone parameter prediction.

[0091] Based on the parameter change matrix, obtain the parameter change value for each parameter and denote it as cbh1(i,j) to cbh a (i, j);

[0092] The parameter changes at different times are statistically analyzed, and the number of parameter changes is denoted as zs. The number of parameter changes that are not equal to 0, bs, is obtained. Combined with the number of parameter changes, zs, the parameter change frequency is calculated to obtain the parameter change frequency bhp(j).

[0093]

[0094] It should be noted that by calculating the frequency of parameter changes, the change status of the parameter can be judged. The higher the frequency of parameter changes, the greater the fluctuation of the parameter.

[0095] Based on the calculation process of the parameter change frequency, the parameter change frequency of each parameter is denoted as bhp1(j) to bhp. a (j);

[0096] bhp a (j) represents the frequency of parameter change of the a-th parameter under the j-th environment;

[0097] By combining the parameter change value with the parameter change frequency, the stable value of the parameter is calculated, and the stable value of the parameter wdz is obtained.

[0098]

[0099] Among them: wdz aa (j) represents the stable value of the aa-th parameter under the j-th environment; bhp aa (j) represents the frequency of parameter change of the aa-th parameter under the j-th environment; cbh aa (i, j) represents the aa-th UAV parameter in the UAV parameter matrix at time i under the j-th environment; aa∈[1, a].

[0100] It should be noted that by combining the parameter change value with the parameter change frequency to judge the parameter fluctuation, the judgment result is more accurate and avoids the error caused by a single parameter. For example, if a parameter is constantly changing, but the change value is very small, it is not accurate enough to judge it by the parameter change frequency alone. Similarly, if a parameter remains stable, but the single parameter change value is large due to error, it is not accurate enough to judge it by the parameter change value alone. Combining the two methods effectively improves the accuracy of the calculation.

[0101] Judge the parameters based on their stable values:

[0102] The mean of parameter changes is calculated, and the mean of parameter changes is used as the threshold yz.

[0103]

[0104] Among them: yz aa (j) represents the threshold of the aa-th parameter in the j-th environment;

[0105] The stable value of parameter wdz is determined by the threshold yz;

[0106] When wdz ≤ yz, it indicates that the parameters are relatively stable, and the parameters should be saved.

[0107] When wdz > yz, it indicates that the parameters are relatively unstable, and the parameters need to be processed.

[0108] The specific steps for handling parameters that are relatively unstable are as follows:

[0109] For parameters that are unstable, we obtain them, denoted as bwd1(j) to bwd. ab (j); Based on the temporal sequence of the parameters, the parameters at all times within time b are obtained, denoted as bwd1(i,j) to bwd. ab (i, j);

[0110] Please see Figure 2 Transfer bwd1(i,j) to bwd ab (i, j) are processed using the least squares method to obtain its regression function, denoted as FH1(i) to FH. ab (i); where FH ab (i) represents the regression value of the ab-th parameter at time i.

[0111] Parameters that are relatively stable are represented by constants;

[0112] It should be noted that: parameters that are relatively stable have relatively stable values ​​over time b and can be represented by constants. For example, if csy1(i,j) is a relatively stable parameter, then csy1(1,j) = csy1(1,j) to csy1(i,j).

[0113] Integrate all parameters, through F1(i,j) to F a Representing (i, j) yields the dynamic model dmx of the UAV;

[0114] Based on the dynamic model of the UAV, dynamic prediction is performed on the UAV to obtain the predicted value dmx(i,j). The predicted value dmx(i,j) is combined with the environmental parameter matrix, and the flight state of the UAV is estimated by optimal linear recursion through Kalman filtering to obtain the optimal parameter matrix of the UAV flight state.

[0115] It should be noted that the Kalman filter is an optimal linear recursive estimation algorithm that combines a system dynamic model with observational data to provide an optimal estimate of the state of a dynamic system under noise interference. Its core idea is to dynamically adjust the estimate of the system state through a prediction-update loop, minimizing the estimation error.

[0116] Monitoring and control module: Acquires real-time data of the UAV, calculates the deviation between the real-time data and the optimal parameter matrix to obtain the deviation value of the UAV, and makes control judgments on the UAV based on the deviation value to determine the operating status of the UAV.

[0117] Real-time data of the UAV is acquired through multi-dimensional sensors. The real-time data of the UAV is judged based on the optimal parameter matrix. The deviation values ​​of the real-time data of the UAV from the optimal parameter matrix at the previous h time points and the deviation values ​​of the real-time data of the UAV from the optimal parameter matrix at the current time point are calculated respectively. The adaptive decision of the UAV is judged based on the change of the deviation value, and the operating state of the UAV is selected based on the judgment result.

[0118] The calculation process for the deviation value is as follows:

[0119] Obtain the optimal parameter matrix zuy for the UAV's flight state. Let us denote the number of parameters in the optimal parameter matrix as us, and the parameters in the optimal parameter matrix as zss; zuy = [zss1, zss2, ..., zss us ];

[0120] The real-time parameters of the UAV are acquired and denoted as ssc; the real-time parameters of the UAV are statistically analyzed to obtain the real-time parameter matrix scj; the data format of the real-time parameter matrix of the UAV is made consistent with that of the optimal parameter matrix to obtain the real-time parameter matrix scj = [ssc1, ssc2, ..., ssc us ];

[0121] The correlation values ​​of the matrix parameters in the real-time parameter matrix of the UAV are determined, and the parameters in the real-time parameter matrix are iteratively checked from ssc1 to ssc. us The parameters are adjusted sequentially, and the correlation values ​​glz1 to glz of the matrix parameters are obtained based on the number of other parameters that change simultaneously when a parameter is adjusted. us ;

[0122] Based on the correlation values, the associated matrix parameters in the real-time parameter matrix are obtained to obtain the correlation parameters; denoted as xgl1 to xgl xg ; Combining the optimal matrix parameters, for xgl1 to xgl xg Obtain the corresponding parameters to get the optimal corresponding parameters; denoted as xzy1 to xzy xg ;

[0123] Based on the correlation parameters and the optimal corresponding parameters, the influence value of each parameter is calculated to obtain the influence value yxz. u ;

[0124]

[0125] Among them: glz u This represents the associated value of the u-th matrix parameter; xgl v xgl represents the v-th associated parameter. v For the vth Optimal corresponding parameters;

[0126] It should be noted that: by performing overall calculations using associated parameters, the accuracy of the influence value calculation is enhanced; by adding 1 to integrate the overall change values ​​of associated parameters, and then increasing them by exponentiation in conjunction with the number of associated parameters, the more associated parameters there are, the greater their influence value.

[0127] The influence ratio of each parameter in the parameter matrix is ​​evaluated by the influence value, and the deviation value is weighted according to the influence value to enhance the accuracy of the deviation value calculation.

[0128] Based on the real-time parameter matrix of the UAV [ssc1, ssc2, ..., ssc us ] and the optimal parameter matrix [zss1, zss2, ..., zss us The parameters in the matrix are calculated, and the influence values ​​y, x, and z of the matrix parameters are combined. u The deviation value of the drone is calculated to obtain the deviation value plz;

[0129]

[0130] Among them: ssc u zss represents the u-th parameter in the real-time parameter matrix. u This represents the u-th parameter in the optimal parameter matrix;

[0131] It should be noted that by assigning weights to parameters through the correlation values ​​of matrix parameters, parameters with a large impact range but small data changes can be more prominent in the calculation, thereby enhancing the accuracy of data calculation;

[0132] The deviation values ​​of the previous h times are calculated according to the formula for calculating the deviation value plz, and plz(1) to plz(h) are obtained.

[0133] The mean deviation value plj is obtained by calculating the mean deviation value of the deviation values ​​of the first h time points;

[0134]

[0135] The mean change of the deviation value over the first h time steps is calculated to obtain the mean change value bhj;

[0136]

[0137] Based on the deviation from the mean plj and the change mean bhj, the current deviation value is used to determine the state and obtain the current state value ztz.

[0138]

[0139] Where: plz(h) represents the deviation value of the previous time step at the current time step;

[0140] It should be noted that: the direction of change of the value is determined by the ratio of the change value to the absolute value of the change value at the current moment; the magnitude of change is determined by the difference between the change value and the mean change value; and the state value is constrained by combining the deviation value and the deviation from the mean to prevent the deviation value from being too large.

[0141] By visually representing the current state of the drone through status values, the deviation of the drone and the trend of deviation can be reflected more directly; by comparing and analyzing the status values, the subsequent operation of the drone can be determined, reducing manual control.

[0142] If the state value ztz is less than 0, it indicates that the parameters of the UAV are far from the optimal parameter matrix and manual intervention is required.

[0143] If the state value ztz is greater than or equal to 0, it indicates that the parameters of the UAV are consistent with the optimal parameter matrix, or that the parameters of the UAV are changing toward the optimal parameter matrix; this is suitable for UAVs to make adaptive decisions.

[0144] Adaptive Decision Module: The adaptive decision module determines the operating status of the UAV, adaptively adjusts the real-time data of the UAV, judges the adaptive adjustment results by combining the optimal parameter matrix, and stores and records the adaptive adjustment results.

[0145] Once it is determined that the UAV can make adaptive decisions, based on the UAV's real-time data and the optimal parameter matrix, the UAV makes adaptive decision judgments to fit the UAV's adaptive decision results with the optimal parameter matrix, and the various parameters of the UAV's operating state tend to the optimal parameter matrix; different UAV adaptive adjustment results are saved, and the saved data is transmitted during the next UAV adaptive adjustment, and the UAV makes adaptive decisions based on the existing adaptive adjustment data.

[0146] Example 2

[0147] Please see Figure 3 The UAV flight control methods based on multi-source sensor fusion and adaptive decision-making include:

[0148] Step S1: Data acquisition module: acquires flight mission information, uses the UAV based on the flight mission information, acquires the UAV's operating parameters and external environment, and obtains UAV information;

[0149] Step S2: Based on the UAV information, obtain the environmental parameter matrix, process the environmental parameter matrix to obtain the UAV dynamic model; use the UAV dynamic model to predict the parameters of the UAV and obtain the predicted values; obtain the UAV parameter matrix at different times through the environmental parameter matrix; perform optimal linear recursive estimation between the UAV parameter matrix at different times and the predicted values ​​to obtain the optimal parameter matrix of the UAV.

[0150] Step S3: Acquire real-time data of the UAV, calculate the deviation between the real-time data of the UAV and the optimal parameter matrix to obtain the deviation value of the UAV, and make control judgments on the UAV based on the deviation value to determine the operating status of the UAV.

[0151] Step S4: Determine the operating status of the UAV as an adaptive decision module, adaptively adjust the real-time data of the UAV, judge the adaptive adjustment result by combining the optimal parameter matrix, and store and record the adaptive adjustment result.

[0152] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to specific implementations. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A UAV flight control system based on multi-source sensor fusion and adaptive decision-making, characterized in that, include: Data acquisition module: Acquires flight mission information, uses the UAV based on the flight mission information, acquires the UAV's operating parameters and external environment, and obtains UAV information; Data processing module: Based on UAV information, obtain the environmental parameter matrix, process the environmental parameter matrix to obtain the UAV dynamic model; use the UAV dynamic model to predict UAV parameters and obtain predicted values; obtain the UAV parameter matrix at different times through the environmental parameter matrix; perform optimal linear recursive estimation between the UAV parameter matrix at different times and the predicted values ​​to obtain the optimal parameter matrix of the UAV. Monitoring and control module: Acquires real-time data of the UAV, calculates the deviation between the real-time data and the optimal parameter matrix to obtain the deviation value of the UAV, and makes control judgments on the UAV based on the deviation value to determine the operating status of the UAV. Adaptive Decision Module: The adaptive decision module determines the operating status of the UAV, adaptively adjusts the real-time data of the UAV, judges the adaptive adjustment results by combining the optimal parameter matrix, and stores and records the adaptive adjustment results.

2. The UAV flight control system based on multi-source sensor fusion and adaptive decision-making according to claim 1, characterized in that, The operating parameters and external environment of the drone are acquired, as follows: Obtain the flight origin, construct a three-dimensional coordinate system from the flight origin, and represent points on the three-dimensional coordinate system through the x, y, and z axes; measure the UAV parameters during UAV movement, specifically including: The flight deflection angle and angular velocity of the UAV, the acceleration of the UAV in the x, y, and z axes, the magnetic force of the UAV on the UAV in the x, y, and z axes, and the position parameters of the UAV are measured. The UAV parameters are statistically analyzed to obtain the UAV parameter matrix cs, cs = [csy1, csy2, ..., csy...]. a ]; where 'a' represents the number of UAV parameters, and 'csy' represents the number of parameters. a This represents the parameters of the a-th UAV; Based on the temporal sequence of UAV parameter measurements, the UAV parameter matrices at different times are obtained and denoted as cs(1) to cs(b). The external environment in which the drone is located is detected, and the type of external environment in which the drone is located is obtained, denoted as c. Based on the type of external environment in which the drone is located, the parameter matrix of the drone is integrated to obtain the environmental parameter matrix hcj.

3. The UAV flight control system based on multi-source sensor fusion and adaptive decision-making according to claim 2, characterized in that, The environmental parameter matrix is ​​processed as follows: Based on the environmental parameter matrix, the parameter matrices of UAVs under the same environment are obtained to obtain the UAV's time-series parameter matrix; based on the UAV's time-series parameter matrix, the UAV parameter matrix at different times is obtained as cs(i,j)=[csy1(i,j), csy2(i,j), ..., csy a [(i, j)]; The difference between the UAV parameters at adjacent different times is calculated to obtain the parameter change matrix cbh; Based on the parameter change matrix, obtain the parameter change value for each parameter and denote it as cbh1(i,j) to cbh a (i, j); The parameter changes at different times are statistically analyzed, and the number of parameter changes is denoted as zs. The number of parameter changes that are not equal to 0, bs, is obtained. Combined with the number of parameter changes, zs, the parameter change frequency is calculated to obtain the parameter change frequency bhp(j). Based on the calculation process of the parameter change frequency, the parameter change frequency of each parameter is denoted as bhp1(j) to bhp. a (j); bhp a (j) represents the frequency of parameter change of the a-th parameter under the j-th environment; By combining the parameter change value with the parameter change frequency, the stable value of the parameter is calculated to obtain the stable value wdz; the parameter is judged based on the stable value.

4. The UAV flight control system based on multi-source sensor fusion and adaptive decision-making according to claim 3, characterized in that, The parameters are judged based on their stable values, as follows: The mean of parameter changes is calculated, and the mean of parameter changes is used as the threshold yz. The stable value of parameter wdz is determined by the threshold yz; When wdz ≤ yz, it indicates that the parameters are relatively stable, and the parameters should be saved. When wdz > yz, it indicates that the parameters are relatively unstable, and the parameters need to be processed.

5. The UAV flight control system based on multi-source sensor fusion and adaptive decision-making according to claim 4, characterized in that, The parameters are processed as follows: For parameters that are unstable, we obtain them, denoted as bwd1(j) to bwd. ab (j); Based on the temporal sequence of the parameters, the parameters at all times within time b are obtained, denoted as bwd1(i,j) to bwd. ab (i, j); Transfer bwd1(i,j) to bwd ab Processing (i, j) yields its regression function, denoted as FH1(i) to FH1(i). ab (i); where FH ab (i) represents the regression value of the ab-th parameter at time i; Parameters that are relatively stable are represented by constants; Integrate all parameters, through F1(i,j) to F a Representing (i, j) yields the dynamic model dmx of the UAV; Based on the dynamic model of the UAV, dynamic prediction is performed on the UAV to obtain the predicted value dmx(i,j). The predicted value dmx(i,j) is combined with the environmental parameter matrix to perform optimal linear recursive estimation of the UAV's flight state, and the optimal parameter matrix of the UAV's flight state is obtained.

6. The UAV flight control system based on multi-source sensor fusion and adaptive decision-making according to claim 1, characterized in that, The deviation between the drone's real-time data and the optimal parameter matrix is ​​calculated, as follows: Obtain the optimal parameter matrix zuy for the UAV's flight state. Let us denote the number of parameters in the optimal parameter matrix as us, and the parameters in the optimal parameter matrix as zss; zuy = [zss1, zss2, ..., zss us ]; The real-time parameters of the UAV are acquired and denoted as ssc; the real-time parameters of the UAV are statistically analyzed to obtain the real-time parameter matrix scj; the data format of the real-time parameter matrix of the UAV is made consistent with that of the optimal parameter matrix to obtain the real-time parameter matrix scj = [ssc1, ssc2, ..., ssc us ]; The correlation values ​​of the matrix parameters in the real-time parameter matrix of the UAV are determined, and the parameters in the real-time parameter matrix are iteratively checked from ssc1 to ssc. us The parameters are adjusted sequentially, and the correlation values ​​glz1 to glz of the matrix parameters are obtained based on the number of other parameters that change simultaneously when a parameter is adjusted. us ; Based on the correlation values, the associated matrix parameters in the real-time parameter matrix are obtained to obtain the correlation parameters; denoted as xgl1 to xgl xg ; Combining the optimal matrix parameters, for xgl1 to xgl xg Obtain the corresponding parameters to get the optimal corresponding parameters; denoted as xzy1 to xzy xg ; Based on the correlation parameters and the optimal corresponding parameters, the influence value of each parameter is calculated to obtain the influence value yxz. u ; Among them: glz u This represents the associated value of the u-th matrix parameter; xgl v xgl represents the v-th associated parameter. v For the vth Optimal corresponding parameters; The deviation value is calculated by using the influence value as a weight and combining the real-time data of the UAV with the optimal parameter matrix.

7. The UAV flight control system based on multi-source sensor fusion and adaptive decision-making according to claim 6, characterized in that, The deviation value is calculated as follows: Based on the real-time parameter matrix of the UAV [ssc1, ssc2, ..., ssc us ] and the optimal parameter matrix [zss1, zss2, ..., zss us The parameters in the matrix are calculated, and the influence values ​​y, x, and z of the matrix parameters are combined. u The deviation value of the drone is calculated to obtain the deviation value plz; Among them: ssc u zss represents the u-th parameter in the real-time parameter matrix. u This represents the u-th parameter in the optimal parameter matrix; The deviation values ​​of the previous h times are calculated according to the formula of the deviation value plz, and plz(1) to plz(h) are obtained; the state of the UAV is judged by the deviation values ​​plz(1) to plz(h).

8. The UAV flight control system based on multi-source sensor fusion and adaptive decision-making according to claim 7, characterized in that, The drone's status is assessed as follows: The mean deviation value plj is obtained by calculating the mean deviation value of the deviation values ​​of the first h time points; The mean change of the deviation value over the first h time steps is calculated to obtain the mean change value bhj; Based on the deviation from the mean plj and the change mean bhj, the current deviation value is used to determine the state and obtain the current state value ztz. Where: plz(h) represents the deviation value of the previous time step at the current time step; If the state value ztz is less than 0, it indicates that the parameters of the UAV are far from the optimal parameter matrix and manual intervention is required. If the state value ztz is greater than or equal to 0, it indicates that the parameters of the UAV are consistent with the optimal parameter matrix, or that the parameters of the UAV are changing toward the optimal parameter matrix; this is suitable for UAVs to make adaptive decisions.

9. The UAV flight control system based on multi-source sensor fusion and adaptive decision-making according to claim 1, characterized in that, The drone's real-time data is adaptively adjusted as follows: Once it is determined that the UAV can make adaptive decisions, based on the UAV's real-time data and the optimal parameter matrix, the UAV makes adaptive decision judgments to fit the UAV's adaptive decision results with the optimal parameter matrix, and the various parameters of the UAV's operating state tend to the optimal parameter matrix; different UAV adaptive adjustment results are saved, and the saved data is transmitted during the next UAV adaptive adjustment, and the UAV makes adaptive decisions based on the existing adaptive adjustment data.

10. A UAV flight control method based on multi-source sensor fusion and adaptive decision-making, applicable to the UAV flight control system based on multi-source sensor fusion and adaptive decision-making as described in any one of claims 1-9, characterized in that, Control methods include: Step S1: Data acquisition module: acquires flight mission information, uses the UAV based on the flight mission information, acquires the UAV's operating parameters and external environment, and obtains UAV information; Step S2: Based on the UAV information, obtain the environmental parameter matrix, process the environmental parameter matrix to obtain the UAV dynamic model; use the UAV dynamic model to predict the parameters of the UAV and obtain the predicted values; obtain the UAV parameter matrix at different times through the environmental parameter matrix; perform optimal linear recursive estimation between the UAV parameter matrix at different times and the predicted values ​​to obtain the optimal parameter matrix of the UAV. Step S3: Acquire real-time data of the UAV, calculate the deviation between the real-time data of the UAV and the optimal parameter matrix to obtain the deviation value of the UAV, and make control judgments on the UAV based on the deviation value to determine the operating status of the UAV. Step S4: Determine the operating status of the UAV as an adaptive decision module, adaptively adjust the real-time data of the UAV, judge the adaptive adjustment result by combining the optimal parameter matrix, and store and record the adaptive adjustment result.

Images