A UAV data transmission method based on state estimation rule

Abstract

The present invention discloses a method for transmitting data of a UAV based on a state estimation rule. The method establishes a state space model in combination with the state data of the communication process of a relay UAV; constructs a state estimation model of the UAV 1 with respect to the target actuator; quantifies a mathematical model of the energy available to two relay UAVs; establishes a communication model between the relay UAV and a ground base station; proposes a mathematical model for optimizing the UAV data transmission strategy based on the state estimation error of the ground base station in establishing the communication model between the relay UAV and the ground base station; analyzes the mathematical model for optimizing the UAV data transmission strategy to obtain the event-triggered data transmission strategy of the UAV; and verifies the event-triggered data transmission strategy of the UAV. The present invention establishes an actuator state observation error analysis model for the communication process of a multi-hop relay UAV, and proposes an event-triggered optimal data transmission strategy based on a dynamic programming method, which is suitable for improving the information transmission security performance of the UAV.

Description

Technical Field

[0001] The present invention relates to the field of automation technology, and in particular to a method for transmitting data of an unmanned aerial vehicle based on a state estimation rule. Background Art

[0002] Traditional ground communication systems mainly rely on ground communication infrastructure such as base stations. However, in emergency communication scenarios, ground users cannot obtain communication services. By using drone-assisted communication, temporary communication services can be provided to ground users, and there are some obvious advantages. First, drones are highly flexible, making the deployment of drone-assisted communication systems faster and more effective in responding to emergencies, and suitable for specific scenarios such as disaster areas. Secondly, based on the mobility of drones, the transmission path can be dynamically adjusted to provide better communication quality for data transmission and reduce transmission interference. In wireless communications, drone relay networks are often used to expand communication range, realize emergency communications, data transmission, etc. In addition, in military communications, wireless communication is often used as the main means of communication. Wireless communication is achieved through wireless transmitters and receivers, which can solve the problem of needing to set up communication lines in advance in wired communications. However, the communication distance of wireless communication is limited and it is easily interfered by other signals on the ground, so the reliability is relatively poor. Drones are a very effective aerial communication platform in wireless communications. While meeting the communication requirements of a large coverage area, they are also convenient for networking with ground base stations and users to achieve flexible and mobile communication. With the development and application of drone technology, using drones as carriers of communication transmitters and receivers can effectively increase the communication distance of wireless communications and improve the reliability and anti-interference ability of transmission.

[0003] Although UAV communication networks have many advantages, multi-hop relay networks bring more complex network structures. The wireless channels between UAVs and ground equipment are open, which may lead to higher data transmission packet loss rates in today's increasingly complex communication environment. Therefore, communication security has gradually become a major issue to consider in wireless communication networks. There may be unknown interferers or eavesdroppers in wireless communication networks. In traditional methods, network-layer-based key encryption technology is mainly used to improve the security performance of communication systems. This method can effectively protect the information sent in wireless communications, but in dynamic wireless networks, this encryption technology will greatly increase the system's computational complexity and network delay, and reduce the performance of the entire communication network.

[0004] State estimation plays a fundamental role in control systems, providing an accurate estimate of the current state of the target execution system and providing the necessary information for the controller to achieve effective control and optimization and improve the performance, stability and reliability of the system. Based on state estimation, the study of control problems is to explore how to use the state estimation results to design and implement effective control strategies, which is of great significance for the realization of automation systems and intelligent control. In addition, due to the particularity of the application environment of wireless body area networks, the problem of continuous energy supply faces huge challenges in wireless body area networks. Traditional battery-powered sensors are constrained by fixed energy and run out of energy as the usage time increases, limiting the sustainability of services. In contrast, energy harvesting technology can collect energy from the external environment and convert it into electrical energy, which is regarded as a feasible solution to extend the service life of sensors. However, compared with the relatively clear available transmission energy of battery-powered sensors, the unpredictable characteristics of harvested energy pose new challenges to the design and analysis of relay drone data transmission strategies. Summary of the invention

[0005] In order to solve the above technical problems, the present invention takes the perspective of information layer security technology design into consideration that relay drones can randomly obtain energy from the external environment for data transmission, and proposes a multi-hop relay drone communication system security analysis and data transmission technology based on state observation error, aiming to ensure the security of the wireless communication process and prevent security threats from causing adverse effects on system operation.

[0006] In order to achieve the above object, the technical solution of the present invention is as follows:

[0007] A method for transmitting data from a drone based on a state estimation rule comprises the following steps:

[0008] Step 1: Establish a state space model based on the state data of the relay UAV communication process;

[0009] Step 2: construct a state estimation model of the target actuator of UAV 1;

[0010] Step 3: quantify the mathematical model of energy available to the two relay drones;

[0011] Step 4: Establish a communication model between the relay drone and the ground base station;

[0012] Step 5: Based on the ground base station state estimation error in step 4, a mathematical model for optimizing the UAV data transmission strategy is proposed;

[0013] Step 6: Analyze the mathematical model for optimizing the UAV data transmission strategy in step 5 to obtain the event-triggered data transmission strategy of the UAV;

[0014] Step 7: Verify the event-triggered data transmission strategy of the drone.

[0015] Preferably, the step 1 specifically comprises the following steps:

[0016] Step 1.1 collects the status information of the UAV communication process to describe the actual system;

[0017] Step 1.2 establishes the state space model of the actuator operation process in the following form:

[0018] x t+1 =Ax t +w t ,

[0019] y t =Cx t +v t ,

[0020] in, is the state vector of the target actuator at time t, is the measurement data observed by the sensor carried by UAV 1, It is an unpredictable external disturbance factor. The measurement noise of UAV 1 satisfies is a positive semidefinite matrix, is a positive definite matrix, When k = j, δ kj =1, otherwise δ kj = 0, the symbol ' represents the transpose of a vector or matrix, w k (w j )′ represents the vector w k and vector w j The transposed product, v k (v j )′ represents the vector v k and vector v j The product of the transpose, w k (v j )′ represents the vector w k and vector v j The product of the transpose, the initial state x0 of the target actuator obeys a Gaussian distribution with an expected value of 0 and a variance of ∑0, and does not depend on the noise w t and v t , assuming that the constant matrix A is unstable and (A,C) is detectable, is positive definite, and the matrices A, C, Q, and R represent the system matrices known to the target actuator.

[0021] Preferably, a state estimation model of the drone 1 with respect to the target actuator is constructed as follows:

[0022]

[0023] Its value can be obtained through the designed Kalman filter,

[0024]

[0025]

[0026]

[0027]

[0028]

[0029] in, represents the state prediction value of the filter at time t, represents the state value of the t filter, represents the error covariance of the state prediction value of the filter at time t, P 0,t Represents the state estimation error covariance of the filter at time t. The Kalman filter iteration process starts from start, represents the gain matrix of the Kalman filter at time t, is the unit matrix, A′, C′ represent the transpose of the system matrix A, C respectively, and introduce the operators h(·) and in is a set of n-dimensional semi-positive matrices, h(X) = AXA + Q and g(X) = X - XC'[CXC' + R] -1 CX, [CXC′+R] -1 represents the inverse of the matrix CXC+R, since (A,C) is detectable, is positive definite, so P 0,t Converge to A fixed point in Without loss of generality, assume that the Kalman filter at UAV 1 has reached a steady state, that is,

[0030] Preferably, the mathematical model for quantifying the energy available to the two relay drones is as follows:

[0031] b t+1 =min{b t -a t ∈+μ t+1 ,b max},

[0032] Among them, b t represents the remaining available energy of the drone at time t, a t is the data transmission decision variable of the UAV at time t. If data is sent to the ground base station at time t, then a t =1, otherwise a t = 0, ∈ is the energy consumed for each data transmission, b max is the battery capacity of the sensor carried by the drone, μ t The energy collected by the drone at each moment.

[0033] Preferably, the communication model between the relay drone and the ground base station is established as follows:

[0034]

[0035] in, and are the target actuator states x t In the minimum mean square error estimation between UAV 2 and the ground base station, the corresponding error covariance matrix is ​​recorded as P 1,t and P 2,t , β i,t , i∈[2] is a Bernoulli random variable sequence indicating whether the data packet is in channel c at time t i Lost during transmission, β i,t = 1 if and only if UAV 2 or the ground base station successfully receives the data sent from UAV 1 or UAV 2 at time t and satisfies Pro.(γ i,t =1) = λ i , where Pro. is the abbreviation of probability.

[0036] Preferably, the ground base station state estimation error proposes a mathematical model for optimizing the UAV data transmission strategy. Specifically, given a time range T, the optimization model is as follows:

[0037]

[0038] Preferably, in step 6, the event-triggered data transmission strategy of the drone is obtained as follows:

[0039] (i) Given the remaining energy b at time t t and the collected energy μ t , when the system matrix and When , the optimal transmission strategy of the relay UAV is to transmit data to the ground base station, that is, Among them, the constant matrix and depends on the remaining energy b at time t t And the collected energy μ t ;

[0040] (ii) Given (P 1,t-1 ,P 2,t-1 ) and μ t , the optimal transmission strategy for the relay UAV is based on the residual energy b t The event-triggered strategy is

[0041]

[0042] in, Depends on the state estimation error covariance P at time t 1,t-1 ,P 2,t-1 And the collected energy μ t .

[0043] Preferably, the step 7 specifically comprises the following steps:

[0044] Step 7.1 Construct a mathematical model equivalent to the UAV objective function:

[0045]

[0046] Among them, tr() represents the trace of the matrix, λ1,λ2 represent parameters, is a constant matrix,

[0047] Step 7.2 Design the recursive function J t (·,·): S→R:

[0048] J T+1 (P1,P2,μ,b)=0,

[0049]

[0050] in, represents the energy collected by the drone at the current moment and the next moment, b represents the energy available to the drone, P1, P2 represent the state estimation error covariance matrix of drone 2 and the ground base station, J t+1 represents the recursive function at time t+1, J T+1 represents the recursive function at time T+1. Since only the transmission before time T is considered, let J T+1 =0, S = (s t )=(P 1,t-1 ,P 2,t-1 ,μ t ,b t ),t∈[T], define the function φ(x):=min{x,b max}, satisfying the property 0≤φ(x)-φ(xc)≤c, c is a constant,

[0051] Step 7.3 Design an equivalent form of the recursive function:

[0052]

[0053] Among them, the formula With formula Corresponding to transmission strategy a t =0 and a t =1;

[0054] Step 7.4 Design the auxiliary function according to step 7.3:

[0055]

[0056] Step 7.5: Apply mathematical induction to verify that event-triggered data transmission strategies (i) and (ii) are valid. According to the auxiliary function designed in step 7.4, we only need to verify Φ t (P1, P2) is monotonically increasing with respect to the error covariance matrix P = (P1 + P2), that is, we only need to verify

[0057]

[0058] The monotonicity of The monotonicity of

[0059] Step 7.6 When t = T + 1, according to the recursive function form of step 7.2, the monotonicity of formula (1) obviously holds. For P ≥ P′ (i.e., P2 ≥ P2′, P1 ≤ P1′) and 0 ≤ bb′ ≤ ∈, assuming that when l = T, …, t + 1, Established, among which J l represents the recursive function at time l, J t+1 represents the recursive function at time t+1, h represents the operator h(·),

[0060] At time t, when P≥P′ and b≥∈, b′≥∈, we can get

[0061]

[0062] When b<∈, b′<∈, we can get

[0063]

[0064] When b≥∈,b′<∈, we can get

[0065]

[0066] From the above calculation, we can know that if the optimal data transmission strategy of the relay drone is a *=0, then we can get

[0067] If the optimal data transmission strategy of the relay drone is a * =1, we can get

[0068]

[0069] Since 0≤bb′≤∈, According to assumption (2), the above inequality holds, that is, when l = t, conclusion (2) still holds, verifying that the relay drone event triggers data transmission strategy (i) is valid.

[0070] Based on the above technical scheme, the beneficial effects of the present invention are as follows: a method for transmitting data of a UAV based on a state estimation rule is provided, and a state space model is established in combination with the state data of the communication process of the relay UAV; a state estimation model of the UAV 1 about the target actuator is constructed; a mathematical model of the energy available to two relay UAVs is quantified; a communication model between the relay UAV and the ground base station is established; a mathematical model for optimizing the UAV data transmission strategy is proposed based on the state estimation error of the ground base station in the communication model between the relay UAV and the ground base station; the mathematical model for optimizing the UAV data transmission strategy is analyzed to obtain the event-triggered data transmission strategy of the UAV; and the event-triggered data transmission strategy of the UAV is verified. From the perspective of information layer security technology design, the present invention considers that the relay UAV can randomly obtain energy from the external environment for data transmission, and proposes a multi-hop relay UAV communication security and data transmission technology based on state observation error, which aims to ensure the security of the wireless communication process and prevent security threats from causing adverse effects on system operation. The method helps to improve the control accuracy of the ground base over the target actuator, and can also improve the communication capability of the entire multi-hop relay UAV system. This method fully considers factors such as limited energy of relay drone nodes, uncertain external environment, random packet loss, measurement noise, etc. when modeling, and models the data transmission problem of multi-hop relay network as an optimization problem under conditional constraints. According to the system performance indicator, namely the state estimation error, the objective function of the optimization problem is analyzed and quantified, and a data transmission strategy based on event triggering is proposed. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 is a schematic diagram of airspace relay drone communication in an embodiment;

[0072] Figure 2 The present invention is a flowchart of an implementation method of a UAV data transmission method based on a state estimation rule in an embodiment. DETAILED DESCRIPTION

[0073] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0074] like Figure 1 As shown in the figure, a system consisting of two UAVs in a certain airspace, a target actuator, and a ground base station is taken as the research object. The goal is to improve the communication efficiency of UAV relays. Considering the energy limitation conditions, an event-triggered data transmission technology is proposed to improve the accuracy of the ground base's control over the target actuator so that it can complete the established related tasks. The specific process is as follows Figure 2 shown.

[0075] Step 1: Combine the state data of the relay drone communication process to establish a state space model:

[0076] 1.1 Collect the status information of the drone communication process to describe the actual system:

[0077] Consider the communication formation and operable area of ​​UAVs in a certain airspace. Figure 1As shown in the figure, two drones are scattered at different coordinate positions in the sky. The two drones can communicate with each other, and the communication is directional. For example, drone 1 collects data on the state of the task target actuator through sensors, and drone 2 can receive the state information from drone 1. Then drone 2 sends the received information to the ground base station through the wireless network. The ground base station analyzes the received data and issues control instructions to adjust the state of the target actuator so that it can complete the relevant tasks. However, the data transmission process through the wireless network may face some challenges, such as signal attenuation, noise and other problems, resulting in data packet loss. Packet loss will make it impossible for the ground base station to correctly receive the data of all relay drones, thereby affecting the accurate estimation of the actuator state. The lost data may contain important information about the actuator, resulting in inaccurate or incomplete state estimation results. Packet loss may also cause the update frequency of the ground base station state estimator to decrease, delay to increase, and thus affect the real-time and response performance of the system. In order to ensure the smooth completion of the target actuator task, the two drones need to cooperate with each other to continuously send the actuator operation status data to the ground base. Since the sensors carried by drones are constrained by fixed energy, the sustainability of the service is limited. Even if energy is extracted from the external environment for data transmission by using energy harvesting technology, the unpredictable characteristics of the external environment still face new challenges in the design and analysis of the data transmission strategy of relay drones. Therefore, in order to quantify the energy available to the relay drone at each moment, the external environment state is modeled using Markov chains. In addition, in order to reduce the random packet loss caused by external factors such as extreme weather and electromagnetic interference, the state estimation performance of the target actuator during the data transmission process is considered, and a mathematical model based on the coordinated scheduling optimization of system resources and tasks is established to quantify the state estimation error of the communication system. A relay drone event-triggered data transmission strategy is proposed to ensure the smooth implementation of the communication system mission, improve the accuracy of the target actuator control and the overall mission execution capability.

[0078] 1.2 The UAV collects the state information of the operation process of the executor of a certain airspace mission and establishes the state space model of the executor operation process in the following form:

[0079] x t+1 =Ax t +w t ,

[0080] y t =Cx t +v t ,

[0081] in, is the state vector of the target actuator at time t, is the measurement data observed by the sensor carried by UAV 1, It is an unpredictable external disturbance factor. The measurement noise of UAV 1 satisfies is a semi-positive definite matrix),

[0082] is a positive definite matrix), When k = j, δ kj =1, otherwise δ kj = 0, the symbol ' represents the transpose of a vector or matrix, w k (w j )′ represents the vector w k and vector w j The transposed product, v k (v j )′ represents the vector v k and vector v j The product of the transpose, w k (v j )′ represents the vector w k and vector v j The product of the transpose, the initial state x0 of the target actuator obeys a Gaussian distribution with an expected value of 0 and a variance of ∑0, and does not depend on the noise w t and v t , assuming that the constant matrix A is unstable and (A,C) is detectable, is positive definite, and the matrices A, C, Q, and R represent the system matrices known to the target actuator.

[0083] Step 2: Construct the state estimation model of UAV 1 regarding the target actuator:

[0084] The smart sensor carried by UAV 1 uses the measurement data {y1, y2, ..., y t} to estimate the actuator state x t , and transmit the measured data transmission to the ground base station through UAV 2 (i.e., channels c1 and c2). The minimum mean square error estimate of the actuator state measured by UAV 1 is recorded as The corresponding estimation error covariance matrix is ​​recorded as

[0085]

[0086] Its value can be obtained through the designed Kalman filter,

[0087]

[0088]

[0089]

[0090]

[0091]

[0092] in, represents the state prediction value of the filter at time t, represents the state value of the t filter, represents the error covariance of the state prediction value of the filter at time t, P 0,t Represents the state estimation error covariance of the filter at time t. The Kalman filter iteration process starts from start, represents the gain matrix of the Kalman filter at time t, is the unit matrix, A′, C′ represent the transpose of the system matrix A, C respectively, and introduce the operators h(·) and in is a set of n-dimensional semi-positive definite matrices, h(X) = AXA′ + Q and g(X) = X-XC′ [CXC′ + R] -1 CX, [CXC′+R] -1 represents the inverse of the matrix CXC′+R, since (A,C) is detectable, is positive definite, so P 0,t Converge to A fixed point in Without loss of generality, assume that the Kalman filter at UAV 1 has reached a steady state, that is,

[0093] Step 3: The mathematical model for quantifying the energy available to the two relay drones is as follows:

[0094] b t+1 =min{b t -a t ∈+μ t+1 ,b max},

[0095] Among them, b t represents the remaining available energy of the drone at time t, a t is the data transmission decision variable of the UAV at time t. If data is sent to the ground base station at time t, then a t =1, otherwise a t = 0, ∈ is the energy consumed for each data transmission, b max is the battery capacity of the sensor carried by the drone, μ tis the energy collected by the drone at each moment. Here it is assumed that two drones can share energy through a directional wireless energy transceiver to make up for the difference in energy collected by the nodes due to some random uncertain factors. min{a, b} represents the minimum value of the two numbers a and b.

[0096] Step 4: Establish the communication model between the relay drone and the ground base station as follows:

[0097]

[0098] in, and are the target actuator states x t In the minimum mean square error estimation between UAV 2 and the ground base station, the corresponding error covariance matrix is ​​recorded as P 1,t and P 2,t , γ i,t , i∈[2] is a Bernoulli random variable sequence indicating whether the data packet is in channel c at time t i Lost during transmission, γ i,t = 1 if and only if UAV 2 or the ground base station successfully receives the data sent from UAV 1 or UAV 2 at time t and satisfies Pro.(γ i,t =1) = λ i , probability λ i Depends on the communication channel c i The channel parameters and the energy consumed by the drone each time it sends data. t = 0, the state estimation error covariance matrix P of the ground base station 2，t =h(P 2，t-1 ), when a t =1, the state estimation error covariance matrix P of the ground base station 2，t satisfy

[0099]

[0100] Among them, pro. is the abbreviation of probability.

[0101] Step 5: Based on the ground base station state estimation error in step 4, a mathematical model for optimizing the UAV data transmission strategy is proposed. Specifically, given the time range T, the optimization model is as follows:

[0102]

[0103] Among them, tr(P 2，t ) is the matrix P 2，t The trace of is the sum of the diagonal elements of the matrix, It means taking the minimum value after summing [·].

[0104] Step 6: Analyze the optimization model of step 5 and propose the event-triggered data transmission strategy of drones:

[0105] (i) Given the remaining energy b at time t t and the collected energy μ t , when the system matrix and When , the optimal transmission strategy of the relay UAV is to transmit data to the ground base station, that is, Among them, the constant matrix and depends on the remaining energy b at time t t And the collected energy μ t ;

[0106] (ii) Given (P 1,t-1 ,P 2,t-1 ) and μ t ,The optimal transmission strategy for relay drones is an event-triggered strategy based on the remaining energy BT, that is,

[0107]

[0108] in, Depends on the state estimation error covariance P at time t 1,t-1 ,P 2,t-1 And the collected energy μ t .

[0109] Step 7: The verification process of the optimal data event triggering data transmission strategy of the relay drone is as follows:

[0110] 7.1 Construct a mathematical model equivalent to the UAV objective function:

[0111]

[0112] 7.2 Designing Recursive Functions (Ignore the time subscripts of some quantities temporarily when there is no ambiguity)

[0113]

[0114]

[0115] in, represents the energy collected by the drone at the current moment and the next moment, b represents the energy available to the drone, P1, P2 represent the state estimation error covariance matrix of drone 2 and the ground base station, J t+1 represents the recursive function at time t+1, J T+1 represents the recursive function at time T+1. Since only the transmission before time T is considered, let J T+1 =0, S = (s t)=(P 1,t-1 ,P 2,t-1 ,μ t ,b t ),t∈[T], define the function φ(x):=min{x,b max}, satisfying the property 0≤φ(x)-φ(xc)≤c, where c is a constant.

[0116] 7.3 Design an equivalent form of recursive function:

[0117]

[0118] The two terms in the brackets on the right side of the equation correspond to transmission strategy a t =0 and a t =1 (i.e., no data transmission or data transmission).

[0119] 7.4 Design auxiliary functions according to 7.3:

[0120]

[0121] 7.5 Apply mathematical induction to verify that event-triggered data transmission strategies (i) and (ii) are valid. According to the auxiliary function designed in step 7.4, we only need to verify Φ t (P1, P2) is monotonically increasing with respect to the error covariance matrix P = (P1 + P2), that is, we only need to verify

[0122]

[0123] The monotonicity of The monotonicity.

[0124] 7.6 When t = T + 1, according to the recursive function form of step 7.2, the monotonicity of equation (1) obviously holds. For P ≥ P′ (i.e., P2 ≥ P2′, P1 ≤ P1′) and 0 ≤ bb′ ≤ ∈, assuming that when l = T, …, t + 1, Established, among which J l represents the recursive function at time l, J t+1 represents the recursive function at time t+1, h represents the operator h(·),

[0125] At time t, when P≥P′ and b≥∈, b′≥∈, we can get

[0126]

[0127] When b<∈, b′<∈, we can get

[0128]

[0129] When b≥∈,b′<∈, we can get

[0130]

[0131] From the above calculation, we can know that if the optimal data transmission strategy of the relay drone is a * =0, then we can get

[0132] If the optimal data transmission strategy of the relay drone is a * =1, we can get

[0133]

[0134] Since 0≤bb′≤∈, According to assumption (2), the above inequality holds, that is, when l = t, conclusion (2) still holds, verifying that the relay drone event triggers data transmission strategy (i) is valid.

[0135] The above is only a preferred implementation of a method for transmitting data via a drone based on state estimation rules disclosed in the present invention, and is not intended to limit the protection scope of the embodiments of this specification. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the embodiments of this specification shall be included in the protection scope of the embodiments of this specification.

Claims

1. A method for transmitting data from an unmanned aerial vehicle based on a state estimation rule, characterized in that: The steps include: Step 1: Establish a state space model based on the state data of the relay UAV communication process; Step 2: Construct the state estimation model of UAV 1 regarding the target actuator as follows: Its value is obtained through the designed Kalman filter, in, represents the state prediction value of the filter at time t, represents the state value of the t filter, represents the error covariance of the state prediction value of the filter at time t, P 0,t Represents the state estimation error covariance of the filter at time t. The Kalman filter iteration process starts from start, represents the gain matrix of the Kalman filter at time t, is the unit matrix, A′, C′ represent the transpose of the system matrix A, C respectively, and introduce operators h(·) and g(·): in is a set of n-dimensional semi-positive matrices, h(X) = AXA′ + Q and g(X) = X-XC′ [CXC′ + R] -1 CX, [CXC′+R] -1 represents the inverse of the matrix CXC′+R, since (A,C) is detectable, is positive definite, so P 0,t Converge to A fixed point in Without loss of generality, assume that the Kalman filter at UAV 1 has reached a steady state, that is, Step 3: quantify the mathematical model of energy available to the two relay drones; Step 4: Establish a communication model between the relay drone and the ground base station; Step 5: Based on the ground base station state estimation error in step 4, a mathematical model for optimizing the UAV data transmission strategy is proposed; Step 6: Analyze the mathematical model for optimizing the UAV data transmission strategy in step 5 to obtain the event-triggered data transmission strategy of the UAV; Step 7: Verify the event-triggered data transmission strategy of the drone.

2. The method for transmitting data from a drone based on a state estimation rule according to claim 1, characterized in that: The step 1 specifically comprises the following steps: Step 1.1 collects the status information of the UAV communication process to describe the actual system; Step 1.2 establishes the state space model of the actuator operation process in the following form: x t+1 =Ax t +w t , y t =Cx t +v t , in, is the state vector of the target actuator at time t, is the measurement data observed by the sensor carried by UAV 1, It is an unpredictable external disturbance factor. The measurement noise of UAV 1 satisfies is a positive semidefinite matrix, is a positive definite matrix, When k = j, δ kj =1, otherwise δ kj = 0, the symbol ' represents the transpose of a vector or matrix, w k (w j )′ represents the vector w k and vector w j The transposed product, v k (v j )′ represents the vector v k and vector v j The product of the transpose, w k (v j )′ represents the vector w k and vector v j The product of the transpose, the initial state x0 of the target actuator obeys a Gaussian distribution with an expectation of 0 and a variance of Σ0, and does not depend on the noise w t and v t , assuming that the constant matrix A is unstable and (A,C) is detectable, is positive definite, and the matrices A, C, Q, and R represent the system matrices known to the target actuator.

3. The method for transmitting data of a drone based on state estimation rules according to claim 2, characterized in that: The mathematical model for quantifying the energy available to two relay drones is as follows: b t+1 =min{b t -a t ∈+μ t+1 ,b max }, Among them, b t represents the remaining available energy of the drone at time t, a t is the data transmission decision variable of the UAV at time t. If data is sent to the ground base station at time t, then a t =1, otherwise a t = 0, ∈ is the energy consumed for each data transmission, b max is the battery capacity of the drone carrying the sensor, μ t The energy collected by the drone at each moment.

4. The method for transmitting data from a drone based on a state estimation rule according to claim 3, characterized in that: The communication model between the relay drone and the ground base station is as follows: in, and are the target actuator states x t In the minimum mean square error estimation between UAV 2 and the ground base station, the corresponding error covariance matrix is ​​recorded as P 1,t and P 2,t , γ i,t , i∈[2] is a Bernoulli random variable sequence indicating whether the data packet is in channel c at time t i Lost during transmission, γ i,t = 1 if and only if UAV 2 or the ground base station successfully receives the data sent by UAV 1 or UAV 2 at time t and satisfies Pro.(γ i,t =1) = λ i , where Pro. is the abbreviation of probability.

5. The method for transmitting data of a drone based on state estimation rules according to claim 4, characterized in that: The ground base station state estimation error proposes a mathematical model for optimizing the UAV data transmission strategy. Specifically, given the time range T, the optimization model is as follows:

6. The method for transmitting data from a drone based on a state estimation rule according to claim 5, characterized in that: In step 6, the event-triggered data transmission strategy of the drone is obtained as follows: (i) Given the remaining energy b at time t t and the collected energy μ t , when the system matrix and When , the optimal transmission strategy of the relay UAV is to transmit data to the ground base station, that is, Among them, the constant matrix and depends on the remaining energy b at time t t And the collected energy μ t ; (ii) Given (P 1,t-1 ,P 2,t-1 ) and μ t , the optimal transmission strategy for the relay UAV is based on the remaining energy b t The event-triggered strategy is in, Depends on the state estimation error covariance P at time t 1,t-1 ,P 2,t-1 And the collected energy μ t .

7. The method for transmitting data of a UAV based on state estimation rules according to claim 6, characterized in that: The step 7 specifically includes the following steps: Step 7.1 Construct a mathematical model equivalent to the UAV objective function: Among them, tr() represents the trace of the matrix, λ1,λ2 represent parameters, is a constant matrix, Step 7.2 Design a recursive function Among them, μ, represents the energy collected by the drone at the current moment and the next moment, b represents the energy available to the drone, P1, P2 represent the state estimation error covariance matrix of drone 2 and the ground base station, J t+1 represents the recursive function at time t+1, J T+1 represents the recursive function at time T+1. Since only the transmission before time T is considered, let J T+1 =0, S = (s t )=(P 1,t-1 ,P 2,t-1 ,μ t ,b t ),t∈[T], define the function φ(x):=min{x,b max }, satisfying the property 0≤φ(x)-φ(xc)≤c, c is a constant, Step 7.3 Design an equivalent form of the recursive function: Among them, the formula With formula Corresponding to transmission strategy a t =0 and a t =1; Step 7.4 Design the auxiliary function according to step 7.3: Step 7.5: Apply mathematical induction to verify that event-triggered data transmission strategies (i) and (ii) are valid. According to the auxiliary function designed in step 7.4, we only need to verify Φ t (P1, P2) is monotonically increasing with respect to the error covariance matrix P = (P1 + P2), that is, we only need to verify The monotonicity of The monotonicity of Step 7.6 When t = T + 1, according to the recursive function form of step 7.2, the monotonicity of formula (1) obviously holds. For P ≥ P′ (i.e., P2 ≥ P2′, P1 ≤ P1′) and 0 ≤ bb′ ≤ ∈, assuming that when l = T, ..., t + 1, Established, among which J l represents the recursive function at time l, J t+1 represents the recursive function at time t+1, h represents the operator h(·), At time t, when P≥P′ and b≥∈, b′≥∈, we can get When b<∈, b′<∈, we can get When b≥∈,b′<∈, we can get From the above calculation, we can know that if the optimal data transmission strategy of the relay drone is a * =0, then we can get If the optimal data transmission strategy of the relay drone is a * =1, we can get Since 0≤bb′≤∈, According to assumption (2), the above inequality holds, that is, when l = t, conclusion (2) still holds, verifying that the relay drone event triggers data transmission strategy (i) is valid.

Images