Unmanned aerial vehicle position prediction method and device, electronic equipment and storage medium

Abstract

The invention discloses an unmanned aerial vehicle position prediction method and device, electronic equipment and a storage medium. The unmanned aerial vehicle position prediction method comprises the steps of predicting a prediction state of an unmanned aerial vehicle at a second moment according to a state of the unmanned aerial vehicle at a first moment; the prediction state at least comprises a first prediction position of the unmanned aerial vehicle at a second moment; and performing iteration by taking the first predicted position as an initial value and taking the minimum distance between the predicted path position of the unmanned aerial vehicle and the route as a constraint condition to obtain an optimized second predicted position. When the prediction position of the unmanned aerial vehicle is estimated, optimization iteration is carried out on the prediction position in combination with the route, so that the obtained prediction position is more accurate and higher in precision.

Description

Technical Field

[0001] The present application relates to the field related to unmanned aerial vehicles, and particularly to a method and device for predicting the position of an unmanned aerial vehicle, an electronic device, and a storage medium. Background Art

[0002] In the related art, the research and application of unmanned aerial vehicles (UAVs) are increasing in the civilian field, such as multi-aircraft collaboration, multi-aircraft surveillance, multi-aircraft task allocation, etc. For a UAV to complete the above tasks, it is necessary to know the position information of the UAV at the second moment and for a period of time in the future, and UAV position prediction is an important part of solving the above requirements. For example, in multi-aircraft collision warning, accurately predicting the position information of the UAV will significantly improve the collision alarm accuracy rate. The existing UAV prediction methods mainly use high-cost sensors such as radar to observe the UAV, and use historical sensor information to predict the future position of the UAV. The available data is single, and the prediction accuracy is poor. Summary of the Invention

[0003] In view of this, embodiments of the present application provide a method and device for predicting the position of an unmanned aerial vehicle, an electronic device, and a storage medium, aiming to obtain a more accurate predicted position of the unmanned aerial vehicle.

[0004] The technical solution of the embodiments of the present application is implemented as follows:

[0005] In a first aspect, a method for predicting the position of an unmanned aerial vehicle is provided, and the method includes:

[0006] Predict the predicted state of the unmanned aerial vehicle at the second moment according to the state of the unmanned aerial vehicle at the first moment; the predicted state at least includes the first predicted position of the unmanned aerial vehicle at the second moment;

[0007] Iterate with the first predicted position as the initial value and with the constraint that the distance between the predicted path position of the unmanned aerial vehicle and the flight path is the smallest to obtain an optimized second predicted position.

[0008] Based on the above solution, the iterating with the first predicted position as the initial value and with the constraint that the distance between the predicted path position of the unmanned aerial vehicle and the flight path is the smallest to obtain an optimized second predicted position includes:

[0009] Construct a flight path equation according to the projection position of the predicted path position on the flight path and multiple known positions on the flight path;

[0010] Based on the predicted path position and the flight path equation, construct an expression for solving the minimum distance between the predicted path position of the unmanned aerial vehicle and the flight path;

[0011] Taking the first predicted position as the initial value, performing least - squares integral iteration based on the expression with the minimum distance between the predicted path position of the unmanned aerial vehicle and the route to obtain the second predicted position.

[0012] Based on the above - mentioned solution, the method further includes:

[0013] Obtaining the first uncertainty of the predicted state according to the motion state of the unmanned aerial vehicle at the first moment;

[0014] Determining the second uncertainty according to the first uncertainty;

[0015] Obtaining the predicted position interval of the unmanned aerial vehicle according to the second uncertainty and the second predicted position.

[0016] Based on the above - mentioned solution, the first uncertainty includes the uncertainty of the first predicted position; and the first uncertainty further includes the uncertainty of the predicted speed of the unmanned aerial vehicle and / or the uncertainty of the predicted acceleration of the unmanned aerial vehicle;

[0017] The determining the second uncertainty according to the first uncertainty includes:

[0018] Based on the uncertainty of the first predicted position and in combination with at least one of the uncertainty of the predicted speed and the uncertainty of the predicted acceleration, determining the second uncertainty.

[0019] Based on the above - mentioned solution, the determining the second uncertainty based on the uncertainty of the first predicted position and in combination with at least one of the uncertainty of the predicted speed and the uncertainty of the predicted acceleration includes: determining the second uncertainty according to at least one of the following functional relationships;

[0020] R e ∝(R p +R v T 2 )

[0021] R e ∝(R p +0.5R a T 4 )

[0022] R e ∝(R p +R v T 2 +0.5R a T 4 );

[0023] The R e is the second uncertainty; the R pis the uncertainty of the first predicted position; the R a is the uncertainty of the predicted acceleration; the R v is the uncertainty of the predicted velocity; the T is the time difference between the first moment and the second moment; the ∝ represents a positive correlation.

[0024] Based on the above solution, determining the second uncertainty according to the first uncertainty includes:

[0025] Determining a scaling factor according to the predicted state of the drone;

[0026] Determining the second uncertainty according to the scaling factor and according to the first uncertainty.

[0027] Based on the above solution, determining the scaling factor according to the predicted state of the drone includes:

[0028] Determining the scaling factor according to the second predicted position corresponding to the predicted state.

[0029] Based on the above solution, the method further includes:

[0030] When the measurement state of the drone is obtained, updating the state of the first moment based on the measurement state.

[0031] In a second aspect, a drone position prediction device is provided, and the device includes:

[0032] A prediction module, configured to predict the predicted state of the drone at a second moment according to the state of the drone at a first moment; the predicted state at least includes the first predicted position of the drone at the second moment;

[0033] An iteration module, configured to perform iteration with the first predicted position as an initial value and with the minimum distance between the predicted path position of the drone and the flight path as a constraint condition to obtain an optimized second predicted position.

[0034] In a third aspect, an electronic device is provided, including: a processor and a memory for storing a computer program that can run on the processor, wherein,

[0035] When the processor is used to run the computer program, it executes the steps of any of the methods in the first aspect.

[0036] In a fourth aspect, a computer storage medium is provided, wherein the storage medium stores instructions, and when the instructions run on an electronic device, the electronic device is enabled to execute any of the technical solutions in the first aspect.

[0037] A fifth aspect provides a computer program product, including a computer program or instructions, characterized in that when the computer program or instructions are executed by a processor, the steps of the method for predicting the position of a drone provided by any technical solution of the first aspect are implemented.

[0038] In the technical solution provided by the embodiments of the present application, when estimating the predicted position of a drone, the predicted position will be optimized and iterated in combination with the flight path, so that the obtained predicted position can be more accurate and have higher precision. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a schematic flowchart of a method for predicting the position of a drone provided by an embodiment of the present application;

[0040] Figure 2 It is a schematic flowchart of a process for obtaining a second predicted position provided by an embodiment of the present application;

[0041] Figure 3 It is a schematic diagram of the flight path of a drone provided by an embodiment of the present application;

[0042] Figure 4 It is a schematic flowchart of a method for predicting the position of a drone provided by an embodiment of the present application;

[0043] Figure 5 It is a schematic structural diagram of a device for predicting the position of a drone provided by an embodiment of the present application;

[0044] Figure 6 It is a schematic structural diagram of an electronic device provided by the embodiments of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0045] The present application will be further described in detail below with reference to the drawings and embodiments.

[0046] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which this application belongs. The terms used in the specification of this application herein are only for the purpose of describing specific embodiments and are not intended to limit this application.

[0047] As Figure 1 shown, the embodiments of the present disclosure provide a method for predicting the position of a drone, and the method includes:

[0048] S1110: Predict the predicted state of the drone at a second moment according to the state of the drone at a first moment; the predicted state at least includes the first predicted position of the drone at the second moment;

[0049] S1120: Iterate with the first predicted position as the initial value and with the constraint that the distance between the predicted path position of the drone and the flight path is minimized, to obtain an optimized second predicted position.

[0050] This drone prediction method can be executed by the on-board terminal of the drone, or by the remote control device of the drone, or by the service platform of the drone.

[0051] In some embodiments, the previous state of the drone at the first moment may include: the position, speed, and / or acceleration of the drone at the first moment. The first moment is the previous moment of the second moment.

[0052] Based on various prediction algorithms, the predicted state of the drone at the second moment can be predicted. The state quantities included in this predicted state are the same as those at the first moment, but the values of the state quantities are different.

[0053] At the initial moment of prediction, the initial state at the initial moment can be obtained according to the sensing data of the positioning sensor and / or various inertial sensors of the drone. Here, the various inertial sensors may include: speed sensors and / or acceleration sensors, etc. In some embodiments, the positioning sensor may include a satellite positioning-based sensor, for example, a sensor for positioning based on the Global Positioning System or the Beidou Satellite System.

[0054] In some embodiments, the Kalman filter is used to predict the predicted state at the second moment based on the previous state at the first moment.

[0055] In some embodiments, the unscented Kalman filter in the Kalman filter is used to predict the current state at the second moment based on the previous state at the first moment.

[0056] Exemplarily, when based on the Kalman rate wave, the predicted state at the second moment can be predicted based on the following functional relationship:

[0057] X k =A k *X k-1 where X k-1 is the state vector of the previous state at the first moment; X k is the state vector of the predicted state at the second moment. A k is the state transition matrix from the first moment to the second moment.

[0058] For example, if the state quantities at the first moment include: position, speed, and acceleration, then the state quantities included in the predicted state also include position, speed, and acceleration.

[0059] In some embodiments, the state quantities at the first moment may at least include position and speed.

[0060] In some embodiments, the state quantity at the first moment may be: the actual state obtained based on the positioning sensor and / or inertial sensor of the unmanned aerial vehicle, or the predicted state at the first moment. For example, when the actual state is obtained, the predicted state at the second moment is predicted using the actual state at the first moment; when the actual state is not obtained, the predicted state at the second moment can be predicted using the predicted state at the first moment.

[0061] The predicted state at least includes a first predicted position.

[0062] There is a certain time difference between the second moment and the first moment. During this time difference, the unmanned aerial vehicle will actually fly through multiple positions continuously based on different flight paths, and these multiple positions are usually the expected passing positions. The expected passing position is one or more positions on a predetermined flight route. If the flight paths of the unmanned aerial vehicle between the first moment and the second moment are different, the expected passing positions are different. In the embodiments of the present disclosure, in order to minimize the yaw of the unmanned aerial vehicle as much as possible and control the unmanned aerial vehicle to fly along the established flight route as much as possible, the first predicted position can be used as the initial value of the iteration, and the minimum distance between the expected passing position on the flight path between the first moment and the second moment and the corresponding position of the heading is used as a constraint condition to optimize the first predicted position to obtain an optimized second predicted position. This second predicted position will be closer to the actual flight position of the unmanned aerial vehicle, thereby improving the accuracy and precision of the position prediction of the unmanned aerial vehicle.

[0063] In some embodiments, as Figure 2 shown, the S1120 may include:

[0064] S1121: Construct a flight route equation according to the projected position of the expected passing position on the flight route and multiple known positions on the flight route;

[0065] S1122: Based on the expected passing position and the flight route equation, construct an expression for solving the minimum distance between the expected passing position of the unmanned aerial vehicle and the flight route;

[0066] S1123: Use the first predicted position as the initial value, and perform least squares integral iteration based on the expression for the minimum distance between the expected passing position of the unmanned aerial vehicle and the flight route to obtain the second predicted position.

[0067] As Figure 3 shown is a flight route. This flight route can be a pre-planned flight line.

[0068] Take multiple points on the flight route, and the positions of these points are known. For example, as Figure 3 shown, points A, B, and C.

[0069] Optionally, at least one of the multiple known positions is at the front end of the projected position of the expected passing position on the route, and at least one point is at the back end of the projected position of the expected passing position on the route.

[0070] Based on the equality of the slopes of the straight lines, the foregoing route equation can be constructed.

[0071] For example, Figure 3 the shown route consists of two broken lines, namely AB and BC, where the coordinates of points A, B, and C are [x a , y a , z a , [x b , y b , z b , [x c , y c , z c , and point P is the projected position of the expected passing position of the UAV, represented by the coordinates [x p , y p , z p . Therefore, Figure 3 the route equation of the route in

[0072]

[0073] is as follows:

[0074]

[0075] d is the distance between the expected passing position and the projected position P.

[0076] This functional relationship can be used to find the minimum distance between the expected passing position and the projected position on the route as a condition for iterative optimization.

[0077] In the above functional relationship, represents the vector formed by points A and P. represents the vector formed by points B and P. represents the vector formed by points B and C. represents the vector formed by points A and B. In the above functional relationship, the symbol × represents the cross product between vectors, and the symbol || represents the modulus of the vector.

[0078] In the embodiments of the present disclosure, the least squares method will be used for optimization iteration, and the initial value of the iteration uses the first predicted position included in the predicted state. Through iteration, a second predicted position can be obtained. This second predicted position is the optimized predicted position, which is a more accurate and higher-precision predicted position than the first predicted position.

[0079] In some embodiments, the method further includes:

[0080] obtaining a first uncertainty of the predicted state according to the motion state of the UAV at the first moment;

[0081] determining a second uncertainty according to the first uncertainty;

[0082] obtaining a predicted position interval of the UAV according to the second uncertainty and the second predicted position.

[0083] In the embodiments of the present disclosure, a first uncertainty will also be predicted when estimating the predicted state. The first uncertainty of the predicted state is obtained from the covariance matrix solved by using the Kalman filter.

[0084] In some embodiments, the following functional relationship may be adopted when determining the first uncertainty based on the Kalman filter:

[0085] where P k represents the first uncertainty at the second moment, and A k is the state transition matrix from the first moment to the second moment; is the transpose of A k . Q is the covariance matrix of the noise. Of course, the above is only an example for determining the first uncertainty, and the specific implementation is not limited to the above example.

[0086] The covariance matrix includes: the uncertainty of the predicted position, the uncertainty of the predicted speed, and the uncertainty of the predicted acceleration.

[0087] In the embodiments of the present disclosure, after determining the first uncertainty of the predicted state, the uncertainty is optimized to obtain a second uncertainty. The second uncertainty reflects the fluctuation range of the predicted position.

[0088] Combining the second uncertainty and the second predicted position to obtain a predicted position interval. This predicted position interval is used as the final predicted position of the UAV. The predicted position interval given indicates the possible range where the UAV will be at the second moment. Using this predicted position interval for anti-collision state processing between UAVs can better avoid collisions between UAVs.

[0089] In some embodiments, the first uncertainty includes the uncertainty of the first predicted position; and the first uncertainty further includes the uncertainty of the predicted speed of the UAV and / or the uncertainty of the predicted acceleration of the UAV.

[0090] In some embodiments, the determining the second uncertainty according to the first uncertainty includes:

[0091] Determine the second uncertainty based on the uncertainty of the first predicted position and in combination with at least one of the uncertainty of the predicted velocity and the uncertainty of the predicted acceleration.

[0092] The second uncertainty is actually the uncertainty of the predicted position. When determining the uncertainty of the predicted position in the present embodiment, not only the uncertainty of the predicted position corresponding to the state at the first moment at the second moment is considered, but also at least one of the predicted velocity and / or the predicted acceleration is considered to determine the second uncertainty. Therefore, the accuracy of the second uncertainty itself can be improved.

[0093] In some embodiments, the determining the second uncertainty based on the uncertainty of the first predicted position and in combination with at least one of the uncertainty of the predicted velocity and the uncertainty of the predicted acceleration includes: determining the second uncertainty according to at least one of the following functional relationships;

[0094] R e ∝(R p +R v T 2 )

[0095] R e ∝(R p +0.5R a T 4 )

[0096] R e ∝(R p +R v T 2 +0.5R a T 4 );

[0097] The R e is the second uncertainty; the R p is the uncertainty of the first predicted position; the R a is the uncertainty of the predicted acceleration; the R v is the uncertainty of the predicted velocity; the T is the time difference between the first moment and the second moment; the ∝ represents a positive correlation.

[0098] In some embodiments, the second uncertainty can be determined according to at least one of the following functional relationships:

[0099] R e =(R p +R v T 2 )

[0100] R e =(Rp +0.5R a T 4 )

[0101] R e = (R p +R v T 2 +0.5R a T 4 ).

[0102] In some embodiments, determining the second uncertainty according to the first uncertainty includes:

[0103] Determining a scaling factor according to the predicted state of the drone;

[0104] Determining the second uncertainty according to the scaling factor and according to the first uncertainty.

[0105] In the embodiments of the present disclosure, the scaling factor will be dynamically determined according to the predicted state of the drone. A scaling factor is determined according to the predicted state of the drone in each prediction period. The scaling factor and the first uncertainty jointly determine the second uncertainty.

[0106] In some embodiments, a function for determining the scaling factor is selected. For example, the scaling factor is determined according to the sigmoid function. For example, the following functional relationship is adopted:

[0107] k = f(f0,th)

[0108] th is a threshold value, and this threshold value can be a preset constant. f is the selected function, and the sigmoid function is commonly used.

[0109] In some embodiments,

[0110] In some embodiments, at this time, the second uncertainty can be determined by at least one of the following functional relationships:

[0111] R e = k * (R p +R v T 2 )

[0112] R e = k * (R p +0.5R a T 4 )

[0113] R e = k * (R p +R v T 2 +0.5R aT 4 )。

[0114] In some embodiments, the method further includes:

[0115] When the measurement state of the UAV is obtained, update the state at the first moment based on the measurement state.

[0116] There may be a certain difference between the predicted state at the first moment and the actual state of the UAV, and the actual state of the UAV can be obtained through measurement. Thus, when the measurement state of the UAV is obtained, the state at the first moment is updated according to the measurement state of the UAV. Thus, when estimating the predicted state at the second moment based on the state at the first moment, a more accurate estimated value can be obtained.

[0117] In some embodiments, the following functional relationship can be used to update the state at the first moment.

[0118]

[0119] X' k-1 = X k + K k (Z k-1 - H k-1 X k-1 )

[0120] P’ k-1 = (I - K k1 * H k-1 ) * P k-1

[0121] K k is the Kalman gain. H k is the transfer matrix from the predicted state corresponding to the first moment to the measurement state. is the transpose of H k . Z k is the measurement state. X' k-1 is the updated state at the first moment. P’ k-1 is the updated state at the first moment. I is the identity matrix.

[0122] After obtaining the updated X' k-1 , the state at the second moment can be predicted based on the updated X' k-1 to obtain the predicted state at the second moment.

[0123] Embodiments of the present disclosure propose a UAV position prediction method based on prior route information,

[0124] First, assume that the acceleration of the UAV follows the Singer model. Based on this, establish a prediction equation model for the position-velocity-acceleration of the UAV, and establish an observation equation with the position and velocity as the observation information.

[0125] Secondly, introduce the prior route information as the constraint equation of the Kalman filter, and obtain the least-squares optimal solution set under this constraint equation. This position is used as the mean value of the final predicted position of the UAV.

[0126] Finally, solve the uncertainty of its predicted position according to the Kalman filter result and the corresponding route information. The final predicted position is a certain area range. As Figure 4 shown, the method for predicting the position of a UAV provided by the embodiment of the present disclosure may include:

[0127] The first step: Establishment of the Kalman filter equation.

[0128] The equations of the Kalman filter may include: a state equation and a measurement equation.

[0129] The state equation is used to determine the state of the UAV at different times.

[0130] Define the position of the UAV as x r and the velocity as v r . The three-axis acceleration a t of the UAV is respectively a tx , a ty , a tz . Here, the Singer model is used to describe the three-axis acceleration of the UAV respectively, that is

[0131]

[0132]

[0133]

[0134] where l x , l y , l z respectively represent the reciprocals of the maneuver time constants of the UAV in the x, y, and z directions, and w tx , w ty , w tz respectively represent Gaussian white noise with a mean of zero. respectively represent the reciprocals of a tx , a ty , a tz .

[0135] That is, the simple motion equation of the UAV is as follows:

[0136]

[0137] where \(l\) is \(l\) x , \(l\) y , \(l\) z that forms a diagonal matrix, and \(w\) t is \(w\) tx , \(w\) ty , \(w\) tz that forms a column vector. respectively represent the reciprocals of \(x\) r , \(v\) r , \(a\) t .

[0138] The measurement equation, which can be used to determine the measurement state of the UAV.

[0139] Since the UAV itself is equipped with a positioning system, the positioning system outputs the measurement state of the UAV at each moment. The measurement state includes: the measurement position, measurement speed, and / or measurement acceleration of the UAV. Therefore, the observation equation is established as follows:

[0140] \(y = Hx + p\)

[0141] where \(x\) is the measurement state composed of the measurement position \(x\) r , measurement speed \(v\) r and measurement acceleration \(a\) t . The measurement state can be a column vector, that is:

[0142]

[0143]

[0144] where,[[]]END]] respectively represent the transposes of \(x\) r , \(v\) r and \(a\) t .

[0145] \(p\) is a column vector composed of the measurement position and speed accuracy output by the UAV positioning system. \(y\) is the measurement position of the UAV.

[0146] Step 2: Predict the UAV position information and uncertainty.

[0147] Estimate the UAV position information and establish the uncertainty using the Kalman filter. The Kalman filter is a recursive mathematical algorithm that approximates the true value by continuously updating the estimated value, combining new measurement data and the previous estimated value. The following are the general steps for predicting the UAV position using the Kalman filter:

[0148] Define the system model: First, establish the state equation and measurement equation of the UAV. Here, refer to Step 1.

[0149] Initialization: Initialize the state estimate and covariance matrix of the Kalman filter. Usually, the initial state can be obtained through sensors such as GPS, and the initial covariance matrix can be set to a relatively large value.

[0150] Prediction stage: At each time step, use the motion model to predict the current state. This results in a predicted state estimate and covariance matrix.

[0151] Update stage: When new measurement data (such as from sensors like GPS, IMU, etc.) is available, compare the predicted state with the measurement data and calculate the Kalman gain. The Kalman gain represents the trade-off between prediction and measurement. Then, use the Kalman gain to update the predicted state and covariance matrix to obtain a more accurate position estimate.

[0152] Repeat prediction and update: In subsequent time steps, continuously repeat the prediction and update process to gradually improve the accuracy of the position estimate. If there is no measurement information in the subsequent stage, keep repeating the prediction process to obtain the predicted position and the predicted covariance matrix. This covariance matrix represents the aforementioned uncertainty.

[0153] Step 3: Introduce prior route information and solve the least squares optimal solution.

[0154] Since drones usually fly according to a pre-planned route, the following constraint conditions can be obtained:

[0155] Reference Figure 3 the route of, the expression of the preset conditions can be obtained:

[0156]

[0157] Let P be the predicted position in the predicted state obtained based on the Kalman filter as the initial value of the iteration, and use it as the initial iteration value for solving the minimum equation d, and then solve the least squares optimal solution.

[0158] During the prediction process, use the above least squares optimal solution for integral iteration to obtain preliminary prediction information, and then also use it as the initial iteration value for solving the minimum equation d, and then solve the least squares optimal solution as the final predicted output expected value, which is denoted as This expected value is the optimized second predicted position.

[0159] Since the unscented Kalman filter has the following advantages compared to the traditional Kalman filter:

[0160] Nonlinear system applicability: The Unscented Kalman Filter (UKF) is designed for nonlinear systems and can thus handle a wider range of system models. Compared with the Kalman Filter (KF), the UKF is more suitable for systems with nonlinear dynamic or observation equations. Both KF and UKF can be used for the estimation of the predicted state in this application, but the UKF is preferably used for the trajectory of the predicted state. The UKF avoids this linearization error by using the unscented transformation and is thus generally more accurate. KF requires the calculation of the Jacobian matrix of the state transition and observation models, while the UKF does not, which simplifies the implementation of the algorithm.

[0161] In many nonlinear systems, the UKF usually performs better than the KF because it is closer to the true system dynamics.

[0162] Therefore, the Unscented Kalman Filter is selected to estimate and predict the position of the UAV.

[0163] The specific UKF filtering algorithm is as follows:

[0164] 1) First, construct the sigma point set:

[0165]

[0166] In the formula, P x are the expected value and covariance matrix of the current random vector respectively, refers to the i-th row or the i-th column of the matrix root mean square. k = α 2 (n + λ) - n, where α takes 1e - 4 and λ takes 0. n is the dimension of the state variable. The expected value represents the predicted state in the predicted state. The covariance represents the first uncertainty.

[0167] 2) Substitute the Sigma point set {χ i} into the nonlinear equation h(·) to obtain:

[0168] Y i = h(χ i ) i = 0, 1,..., 2n

[0169] 3) Calculate the mean and variance of y

[0170] The expression is as follows:

[0171]

[0172]

[0173] W i (m)is the first weight of the i-th state variable in the predicted state. W i (c) is the second weight of the i-th state variable in the predicted state. In the above formula, W i (m) is related to W i (c) is calculated as follows:

[0174]

[0175]

[0176] W i (m) = W i (c) = k / [2(n + k)], i = 1, …, 2n

[0177]

[0178]

[0179] where Q k represents the noise covariance matrix.

[0180] If there is updated measurement data, it means that the measurement state has been obtained, and the following update operation is performed.

[0181] where γ represents the measurement state obtained from the measurement data.

[0182]

[0183] where R k represents the covariance matrix of the noise

[0184]

[0185]

[0186]

[0187] K represents the transition matrix for the transition from the measurement state to the transition state.

[0188]

[0189] Since the system is subject to the least squares constraint, after obtaining the preliminary estimation information the least squares algorithm is used to obtain:

[0190]

[0191] The initial value of P is The convergent solution of P is the final estimated information.

[0192] Step 4: Combine the uncertainty to output the final result.

[0193] The above Kalman filter will output the uncertainty of each state, and the uncertainties of its position, velocity, and acceleration are R p , R v , R a respectively. The uncertainty of the entire position after time T0 In the above formula, k is the scaling factor of the uncertainty. It is related to the current predicted mean and the position of the flight path. Its calculation method is as follows:

[0194] k = f(f0, th)

[0195] th is the threshold value. f is the selected function. For example, f can be the sigmoid function.

[0196] Among them, the position coordinates of P are substituted with the aforementioned predicted position .

[0197] Therefore, the final result output by the entire system is the expected value The uncertainty is an interval with R e . Since the values in the three directions of R e are not necessarily equal, the output is an ellipsoidal interval.

[0198] In summary, the technical solution provided by the embodiments of the present disclosure uses the flight path information as the constraint equation of the unscented Kalman filter, obtains the least squares optimal solution set under this constraint equation, and uses this position as the method for the final predicted position mean of the unmanned aerial vehicle. A method for obtaining the uncertainty of the current predicted position by fusing the uncertainties of the position, velocity, and acceleration output by the Kalman filter and the prior flight path, and obtaining the predicted position of the regional range.

[0199] By introducing and fusing prior flight path information, the prediction accuracy is relatively poor. This patent introduces prior UAV flight path information, uses the flight path information as a constraint equation for the unscented Kalman filter, and obtains the least squares optimal solution set under this constraint equation. This position serves as the mean of the final predicted position of the UAV. In this way, the predicted position of the UAV is constrained by the flight path information, thereby improving the prediction position accuracy. Secondly, the uncertainty of the current predicted position is obtained by fusing the uncertainty of the position, velocity, and acceleration output by the Kalman filter with the uncertainty between the prior flight path. The obtained predicted position is a certain area range rather than a specific value. The advantage of this solution is that it allows monitoring personnel to intervene and command the flight more conveniently in advance.

[0200] The method of this embodiment of the present disclosure can be used in the following aspects:

[0201] Safety monitoring: UAV collision warning can be used for safety monitoring, such as monitoring the safety around the UAV and monitoring the UAV trajectory. Through UAV collision warning, relevant departments can monitor the UAV status in real time, discover and handle abnormal situations in a timely manner, and improve the safety of the UAV.

[0202] Risk assessment: UAV collision warning can be used for risk assessment, such as assessing the collision risk of the UAV and evaluating the usage efficiency of the UAV. Through UAV collision warning, relevant departments can discover and handle risks in a timely manner, and improve the usage efficiency and safety of the UAV.

[0203] Insurance claims: UAV collision warning can be used for insurance claims. For example, insurance companies can make claims based on UAV collision warning data. Through UAV collision warning, insurance companies can assess the collision risk of the UAV, formulate corresponding claim policies according to the risk situation, and improve the cost performance and safety of insurance.

[0204] Commercial applications: UAV collision warning can be used for commercial applications, such as monitoring the operation of the UAV and evaluating the performance of the UAV. Through UAV collision warning, enterprises can monitor the operation of the UAV in real time, and improve the safety and efficiency of commercial applications.

[0205] In the embodiments of the present disclosure, the future trajectory of the drone and the collision probability between drones are displayed at the front end. By changing the acceleration of the drone and observing the future trajectory of the drone displayed at the front end, it can be seen whether the trajectory is fused with the drone route information. Especially when the route direction changes, if it is not fused with the route information, the predicted position of the drone will be along the previous route direction, while the predicted position fused with the route information will be along the new route direction; whether the predicted position displayed at the front end is a point or a region; the same flight test is carried out with high-precision sensors and low-precision sensors, and by observing whether the predicted area displayed at the front end changes, it can be verified whether the uncertain calculation of the prediction is combined with the accuracy of the sensors.

[0206] As Figure 5 shown, the embodiments of the present disclosure provide a drone position prediction device, including:

[0207] A prediction module 110, configured to predict the predicted state of the drone at a second moment according to the state of the drone at a first moment; the predicted state at least includes the first predicted position of the drone at the second moment;

[0208] An iteration module 120, configured to perform iteration with the first predicted position as the initial value and with the constraint that the distance between the predicted path position of the drone and the route is the smallest, to obtain an optimized second predicted position.

[0209] In some embodiments, the iteration module is configured to construct a route equation according to the projection position of the predicted path position on the route and multiple known positions on the route; based on the predicted path position and the route equation, construct an expression for solving the minimum distance between the predicted path position of the drone and the route; with the first predicted position as the initial value, perform least-squares integral iteration based on the expression for the minimum distance between the predicted path position of the drone and the route, to obtain the second predicted position.

[0210] In some embodiments, the prediction module is further configured to obtain the first uncertainty of the predicted state according to the motion state of the drone at the first moment.

[0211] The device further includes:

[0212] A determination module, configured to determine a second uncertainty according to the first uncertainty;

[0213] An obtaining module, configured to obtain a predicted position interval of the drone according to the second uncertainty and the second predicted position.

[0214] In some embodiments, the first uncertainty includes the uncertainty of the first predicted position; and the first uncertainty further includes the uncertainty of the predicted speed of the drone and / or the uncertainty of the predicted acceleration of the drone.

[0215] The determining module is configured to determine the second uncertainty based on the uncertainty of the first predicted position and in combination with at least one of the uncertainty of the predicted speed and the uncertainty of the predicted acceleration.

[0216] In some embodiments, the determining module is specifically configured to determine the second uncertainty according to at least one of the following functional relationships;

[0217] R e ∝(R p +R v T 2 )

[0218] R e ∝(R p +0.5R a T 4 )

[0219] Re∝(Rp + RvT 2 + 0.5RaT 4 );wherein the R e is the second uncertainty; the R p is the uncertainty of the first predicted position; the R a is the uncertainty of the predicted acceleration; the R v is the uncertainty of the predicted speed; the T is the time difference between the first moment and the second moment; and the ∝ represents a positive correlation.

[0220] In some embodiments, the determining module is specifically configured to determine a scaling factor according to the predicted state of the drone; and determine the second uncertainty according to the scaling factor and according to the first uncertainty.

[0221] In some embodiments, the determining module is specifically configured to determine the scaling factor according to the second predicted position corresponding to the predicted state.

[0222] In some embodiments, the device further includes:

[0223] An updating module, configured to update the state at the first moment based on the measurement state when the measurement state of the drone is obtained.

[0224] Based on the hardware implementation of the above program modules, and in order to implement the method of the embodiments of the present application, the embodiments of the present application further provide an electronic device. Figure 6Only an exemplary structure of the access control device is shown, rather than all structures, which can be implemented as needed. Figure 6 Some or all of the structures shown.

[0225] As Figure 6 shown, the electronic device 1000 provided by the embodiment of the present application includes: at least one processor 1001, a memory 1002, a user interface 1003, and at least one network interface 1004. Each component in the electronic device 1000 is coupled together through a bus system 1005. It can be understood that the bus system 1005 is used to realize the connection and communication between these components. In addition to including a data bus, the bus system 1005 also includes a power bus, a control bus, and a status signal bus. However, for the sake of clear illustration, in Figure 6 all kinds of buses are labeled as the bus system 1005.

[0226] Among them, the user interface 1003 may include a display, a keyboard, a mouse, a trackball, a click wheel, a button, a touchpad, or a touch screen, etc.

[0227] The memory 1002 in the embodiment of the present application is used to store various types of data to support the operation of the access control device. Examples of these data include: any computer program for operating on the access control device.

[0228] The method for predicting the position of the unmanned aerial vehicle disclosed in the embodiment of the present application can be applied to or implemented by the processor 1001. The processor 1001 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the method for predicting the position of the unmanned aerial vehicle can be completed by the integrated logic circuit in the hardware of the processor 1001 or the instructions in the form of software. The above-mentioned processor 1001 may be a general-purpose processor, a digital signal processor (DSP, Digital Signal Processor), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor 1001 can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiment of the present application. The general-purpose processor may be a microprocessor or any conventional processor, etc. Combining the steps of the method disclosed in the embodiment of the present application, it can be directly embodied as being executed and completed by a hardware decoding processor, or by a combination of hardware and software modules in the decoding processor. The software module may be located in a storage medium, and this storage medium is located in the memory 1002. The processor 1001 reads the information in the memory 1002 and combines its hardware to complete the steps of the method for predicting the position of the unmanned aerial vehicle provided by the embodiment of the present application.

[0229] In an exemplary embodiment, the electronic device may be implemented by one or more application specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field programmable gate arrays (FPGAs), general purpose processors, controllers, microcontroller units (MCUs), microprocessors, or other electronic components for performing the foregoing method.

[0230] It can be understood that the memory 1002 can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM, Read Only Memory), a programmable read-only memory (PROM, Programmable Read-Only Memory), an erasable programmable read-only memory (EPROM, Erasable Programmable Read-Only Memory), an electrically erasable programmable read-only memory (EEPROM, Electrically Erasable Programmable Read-Only Memory), a ferromagnetic random access memory (FRAM, ferromagnetic random access memory), a flash memory (Flash Memory), a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD-ROM, Compact Disc Read-Only Memory); the magnetic surface memory can be a disk memory or a tape memory. The volatile memory can be a random access memory (RAM, Random Access Memory), which is used as an external cache. By way of example but not limitation, many forms of RAM are available, such as a static random access memory (SRAM, Static Random Access Memory), a synchronous static random access memory (SSRAM, Synchronous Static Random Access Memory), a dynamic random access memory (DRAM, Dynamic Random Access Memory), a synchronous dynamic random access memory (SDRAM, Synchronous Dynamic Random Access Memory), a double data rate synchronous dynamic random access memory (DDR SDRAM, Double Data Rate Synchronous Dynamic Random Access Memory), an enhanced synchronous dynamic random access memory (ESDRAM, Enhanced Synchronous Dynamic Random Access Memory), a sync link dynamic random access memory (SLDRAM, SyncLink Dynamic Random Access Memory), a direct rambus random access memory (DRRAM, Direct Rambus Random Access Memory).The memories described in the embodiments of the present application are intended to include, but are not limited to, these and any other suitable types of memories.

[0231] In an exemplary embodiment, the embodiments of the present application also provide a storage medium, namely a computer storage medium, specifically a computer-readable storage medium. For example, it includes a memory 1002 storing a computer program, and the above computer program can be executed by a processor 1001 of an electronic device to complete the steps of the method in the embodiments of the present application. The computer-readable storage medium can be a memory such as ROM, PROM, EPROM, EEPROM, Flash Memory, magnetic surface memory, optical disc, or CD-ROM.

[0232] It should be noted that: First, second, etc. are used to distinguish similar objects and do not necessarily describe a specific order or sequence.

[0233] In addition, the technical solutions described in the embodiments of the present application can be arbitrarily combined without conflict.

[0234] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for predicting the position of an unmanned aerial vehicle, characterized in that, The method includes: Predicting the predicted state of the UAV at the second moment according to the state of the UAV at the first moment; the predicted state at least includes the first predicted position of the UAV at the second moment; Iterating with the first predicted position as the initial value and with the constraint that the distance between the predicted path position of the UAV and the flight path is minimized, to obtain an optimized second predicted position.

2. The method according to claim 1, characterized in that, The iterating with the first predicted position as the initial value and with the constraint that the distance between the predicted path position of the UAV and the flight path is minimized, to obtain an optimized second predicted position, includes: Constructing a flight path equation according to the projected position of the predicted path position on the flight path and multiple known positions on the flight path; Based on the predicted path position and the flight path equation, constructing an expression for solving the minimum distance between the predicted path position of the UAV and the flight path; Taking the first predicted position as the initial value, and performing least-squares integral iteration based on the expression for the minimum distance between the predicted path position of the UAV and the flight path, to obtain the second predicted position.

3. The method according to claim 1 or 2, characterized in that, The method further includes: Obtaining the first uncertainty of the predicted state according to the motion state of the UAV at the first moment; Determining a second uncertainty according to the first uncertainty; Obtaining the predicted position interval of the UAV according to the second uncertainty and the second predicted position.

4. The method according to claim 3, characterized in that, The first uncertainty includes the uncertainty of the first predicted position; and the first uncertainty further includes the uncertainty of the predicted speed of the UAV and / or the uncertainty of the predicted acceleration of the UAV; The determining the second uncertainty according to the first uncertainty includes: Based on the uncertainty of the first predicted position and in combination with at least one of the uncertainty of the predicted speed and the uncertainty of the predicted acceleration, determining the second uncertainty.

5. The method according to claim 4, wherein The determining the second uncertainty based on the uncertainty of the first predicted position and in combination with at least one of the uncertainty of the predicted speed and the uncertainty of the predicted acceleration includes: determining the second uncertainty according to at least one of the following functional relationships; R e ∝(R p +R v T 2 ) R e ∝(R p +0.5R a T 4 ) Re ∝ (Rp + RvT 2 + 0.5RaT 4 ); The R e is the second uncertainty; The R p is the uncertainty of the first predicted position; The R a is the uncertainty of the predicted acceleration; The R v is the uncertainty of the predicted velocity; The T is the time difference between the first moment and the second moment; The ∝ represents a positive correlation.

6. The method according to claim 3, wherein The determining the second uncertainty according to the first uncertainty includes: Determining a scaling factor according to the predicted state of the UAV; Determining the second uncertainty according to the scaling factor and according to the first uncertainty.

7. The method according to claim 6, wherein The determining the scaling factor according to the predicted state of the UAV includes: Determining the scaling factor according to the second predicted position corresponding to the predicted state.

8. The method according to claim 1 or 2, characterized in that The method further includes: When the measurement state of the UAV is obtained, updating the state at the first moment based on the measurement state.

9. A UAV position prediction device, characterized in that, The device includes: A prediction module, configured to predict the predicted state of the UAV at the second moment according to the state of the UAV at the first moment; the predicted state at least includes the first predicted position of the UAV at the second moment; An iteration module, configured to iterate with the first predicted position as the initial value and with the constraint that the distance between the predicted path position of the UAV and the flight path is minimized, to obtain an optimized second predicted position.

10. An electronic device, characterized in that, Comprising: A processor and a memory for storing a computer program that can run on the processor, wherein, The processor is configured to execute the steps of the method according to any one of claims 1 to 8 when running the computer program.

11. A computer storage medium, wherein, The storage medium stores instructions that, when run on an electronic device, cause the electronic device to execute the steps of the method according to any one of claims 1 to 8.

12. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed by the processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images