阅读说明：本技术 一种基于序贯融合算法的多auv协同定位方法 (Multi-AUV (autonomous Underwater vehicle) cooperative positioning method based on sequential fusion algorithm ) 是由 韩春艳 王玮 李雪扬 张颂 于 2021-05-20 设计创作，主要内容包括：本发明公开了一种基于序贯融合算法的多AUV协同定位方法,包括：(1)建立多AUV协同定位系统模型,连续状态方程和离散观测方程；(2)量测信息采用错时采样、错时通信的方式；(3)采用卡尔曼滤波算法,给出单个量测信息的局部分段连续滤波器；(4)采用序贯融合方法,逐步处理采样周期内观测数据,给出当前时刻AUV的位姿信息精确估计值。本发明有益效果：量测丢失过程被描述为对角阵型随机变量,其中分量为Bernoulli随机过程,更具有普适性；采用具有串行结构的序贯融合算法设计分段连续序贯融合滤波器,给出具有高精度和实时性的多AUV协同定位方法,使其包含采样时刻和采样区间的精确估计值。本发明不仅可降低计算负担,还可以提高定位的实时性。(The invention discloses a multi-AUV (autonomous Underwater vehicle) cooperative positioning method based on a sequential fusion algorithm, which comprises the following steps: (1) establishing a multi-AUV cooperative positioning system model, a continuous state equation and a discrete observation equation; (2) the measurement information adopts the modes of time-staggered sampling and time-staggered communication; (3) a local segmented continuous filter of single measurement information is given by adopting a Kalman filtering algorithm; (4) and (3) processing observation data in a sampling period step by adopting a sequential fusion method, and giving an accurate estimation value of the pose information of the AUV at the current moment. The invention has the beneficial effects that: the measurement loss process is described as a diagonal matrix type random variable, wherein the components are Bernoulli random processes and have universality; a sequential fusion algorithm with a serial structure is adopted to design a segmented continuous sequential fusion filter, and a multi-AUV co-location method with high precision and real-time performance is provided, so that the multi-AUV co-location filter comprises accurate estimated values of sampling time and sampling interval. The invention can not only reduce the calculation burden, but also improve the real-time performance of positioning.)

1. A mathematical model of a multi-AUV co-location system in an underwater acoustic communications network, comprising:

wherein X (t) e Rⁿ，U(t)∈R^mRespectively representing the position information vector and the control input vector of an AUV in the underwater acoustic communication system, wherein t is more than or equal to 0;indicates the l th received from AUV_iMeasurement information of individual main AUV, where k_iDenotes the l_iThe time when the main AUV measurement information arrives is k for every two AUV sampling intervals_i+1-k_iFor convenience of presentation, let h be 1; w (t) ε R^pRepresenting system process noise, including velocity measurement noise and orientation measurement noise;representing underwater acoustic distance measurement noise between the AUVs; random variableDescribing the measurement loss situation; a, B, G, H are appropriate constant matrixes respectively.

2. A multi-AUV cooperative positioning method based on a sequential fusion algorithm is characterized by comprising the following steps:

(1) using the mathematical model of claim 1 as a mathematical model of an underwater multi-AUV co-location system;

(2) determining a sampling period (k, k + 1)]The l slave AUV receives the measurement information of m master AUVs in turn in one sampling period, that is

(3) Based on Kalman filtering algorithm, single measurement information is givenThe local filter of (2):

wherein the content of the first and second substances,a state-filter gain array is represented,a filter error variance matrix is represented, satisfying the following equation:

here, the first and second liquid crystal display panels are,S(k_i) Satisfies the following Lyapunov equation:

(4) miningUsing a sequential fusion method, according to the sequence of local measurement information arriving at a fusion estimation center, gradually processing observation data in a sampling period, giving a fusion estimation value in real time, and finally giving accurate estimation position information of the current time, wherein the sequential fusion process comprises the following steps: based on the first arriving₁Individual main AUV measurement informationCan obtain t ═ k₁Temporal local filterAnd based on the time period k₁＜t≤k₂The local predictor obtains t as k₂A predicted value of the time; based on k₂First of arrival of time₂Individual main AUV measurement informationObtaining a local filterAnd so on, based on the measurement information of each main AUV arriving in sequenceCan obtain the first_mA filterWhen t is k +1, the l-th state-optimal sequential fusion filter from the AUV is obtained as

3. The mathematical model of the multi-AUV co-location system in an underwater acoustic communication network as claimed in claim 1, comprising: (1) the measurement information is transmitted by adopting a time-staggered sampling and time-staggered communication mode, so that the first slave AUV has one sampling period(k,k+1]Sequentially receiving the measurement information of m main AUVs, and numbering l according to the sequence₁,...,l_mIn the case of no delay, the ideal discrete time at which the measurement information arrives sequentially is recorded as k₁,k₂,...,k_mThen the relationship is satisfied: k is less than k₁＜k₂＜…＜k_mK + 1; (2) the process of measurement loss is described as a diagonal matrix type random variable, i.e.In which component(s)For the Bernoulli stochastic process, the probability distribution is Meaning that the metrology data was successfully received,meaning that no metrology data is received but a metrology noise term is present in the observation.

4. The multi-AUV co-location method based on the sequential fusion algorithm according to claim 2, wherein: a segmented continuous sequential fusion filter is designed by adopting a sequential fusion algorithm with a serial structure, a multi-AUV (autonomous Underwater vehicle) co-location method with high precision and real-time performance is provided, the method not only comprises an accurate estimation value of a sampling moment, but also comprises an accurate estimation value of a sampling interval, and the fusion process is characterized in that: based on k arriving first₁Local segmented continuous filter designed by time measurement informationFurther with k₂The time arrival measurement information is fused to generate a local segmented continuous filterAnd analogizing in turn, fusing all the measurement information to obtain the estimation of the AUV pose information.

Technical Field

The invention relates to the technical field of network control, and provides a multi-AUV (autonomous underwater vehicle) co-location method of an underwater acoustic communication system with loss measurement, which mainly aims at the underwater acoustic communication system under the influence of a complex underwater environment.

Background

With the increasing of the ocean exploration and resource development, it is increasingly difficult for a single underwater Autonomous operation robot (AUV) to meet the increasingly complex and diverse requirements of human beings on an underwater robot operating system, so that a multi-AUV cooperative operating system receives more and more attention and intensive research of scholars at home and abroad, and becomes a mainstream trend and development direction of the field of underwater robots. The underwater high-precision navigation positioning is a key ring for completing underwater cooperative operation tasks, and plays a decisive role in mutual high-efficiency cooperative operation for autonomous task implementation of underwater operation robots. In view of the many advantages that multi-AUV co-location has not had with each AUV location alone, multi-AUV co-location has become a focus of research in recent years.

However, due to the narrow-band property, the multipath property and the fading property which are peculiar to the complex underwater environment and the underwater acoustic communication network, the underwater acoustic signal is attenuated and even interfered when being transmitted in the underwater acoustic communication channel, and further the phenomenon that the measurement positioning information of the AUV is delayed or lost is caused, which brings great difficulty to the research of the multi-AUV collaborative navigation positioning. In order to achieve underwater positioning with high accuracy and real-time performance, the influence of these random factors needs to be fully considered in the model. However, in the existing positioning algorithm, the influence of some random factors is often selectively ignored. Therefore, the design of the cooperative positioning algorithm capable of improving the positioning accuracy of the multi-AUV navigation has important theoretical significance and practical value.

As is well known, the co-location problem can be attributed to the state estimation problem, so that the filtering algorithm becomes a key factor affecting the co-location accuracy and real-time performance. Research on the filtering algorithm at present has achieved abundant results. The sequential fusion filtering algorithm adopts the current time sensor observation data to gradually update the state estimation value in a sequential processing mode, and finally obtains the current state real-time optimal estimation value. The inventor finds that the purpose of reducing the calculation burden is achieved by the sequential fusion filtering algorithm because the centralized dimension expansion of the observed values is dispersed to be processed according to the sequence of each local observation, so that the solved high-dimensional matrix is converted into the solved low-dimensional matrix. And the method consists of sequentially calculating a plurality of local Kalman filters, so that the global optimality of the estimation precision can be guaranteed. However, the research of designing the multi-AUV co-location algorithm of the underwater acoustic communication system based on the sequential fusion algorithm is relatively deficient.

Disclosure of Invention

In order to solve the problems, the invention provides a multi-AUV (autonomous Underwater vehicle) co-location method based on a sequential fusion algorithm, and aims to provide the multi-AUV co-location method with higher location accuracy and real-time performance aiming at the condition of loss measurement in an underwater acoustic communication network by designing a sequential fusion filter.

In order to achieve the above object, the present invention adopts the following technical solutions. The specific steps of the scheme are as follows.

Step 1: acquiring sensor parameters of an underwater acoustic communication network system, transmitting measurement information in a time-staggered sampling and time-staggered communication mode, and establishing a master-slave multi-AUV state space model:

wherein X (t) e Rⁿ，U(t)∈R^mRespectively representing the position information vector and the control input vector of an AUV in the underwater acoustic communication system, wherein t is more than or equal to 0;indicates the l th received from AUV_iMeasurement information of individual main AUV, where k_iDenotes the l_iThe time when the main AUV measurement information arrives is k for every two AUV sampling intervals_i+1-k_iFor convenience, h is assumed to be 1; w (t) ε R^pRepresenting system process noise, including velocity measurement noise and orientation measurement noise;representing underwater acoustic distance measurement noise between the AUVs; diagonal matrix type random variableDescribing a metrology loss profile, wherein the componentsObeying a Bernoulli distribution with a probability distribution ofi，j＝1，…m。 Meaning that the metrology data was successfully received,meaning that no metrology data is received but a metrology noise term is present in the observation. A, B, G, H are respectively appropriate dimension constant matrixes. Because the measurement information is transmitted by adopting a time error sampling and time error communication mode, the first slave AUV has one sampling period (k, k + 1)]Sequentially receiving the measurement information of m main AUVs, and numbering l according to the sequence₁，l₂，...，l_mWhen the delay is not considered, the ideal discrete time at which the measurement information arrives sequentially is recorded as k₁，k₂，...，k_mThen the relationship is satisfied: k is less than k₁＜k₂＜…＜k_m＝k+1。

Step 2: based on Kalman filtering algorithm, single measurement information is given in consideration of measurement loss situationThe calculation equation of the local filtering algorithm is as follows:

wherein the content of the first and second substances,a state-filter gain array is represented,a filter error variance matrix is represented, satisfying the following equation:

here, the first and second liquid crystal display panels are,S(k_i) Satisfies the following Lyapunov equation:

and step 3: local filtering values (k) according to step 2 based on the measurement information of the m main AUVs_i-1＜t≤k_i) Give the ith slave AUV in one sample period (k, k + 1)]And (3) sequential fusion filtering process:

(1) based on the first arriving₁Individual main AUV measurement informationCan obtain t ═ k₁Temporal local filterAnd based on the time period k₁＜t≤k₂The local predictor obtains t as k₂A predicted value of the time;

(2) based on k₂First of arrival of time₂Individual main AUV measurement informationObtaining a local filterAnd so on, based on the measurement information of each main AUV arriving in sequenceCan obtain the first_mA filter

(3) When t is k +1, the l-th state-optimal sequential fusion filter from the AUV is obtained as

Preferably, in step 1, w (t) is a wiener process with zero mean and variance matrix Q;is a zero mean and variance matrix ofWhite gaussian noise of (1); the initial value X (0) of the state of the system is that the mean value is zero and the variance is V₀And X (0) and w (t),andl_i＝l₁，...，l_mare independent of each other.

Due to the adoption of the technical scheme, the invention has the remarkable technical effects that:

(1) the underwater multi-AUV cooperative positioning technology is researched by taking an underwater acoustic communication network widely applied by military and civil as a background, a state space model with loss measurement is established by considering a plurality of complicated and changeable uncertainties in the actual process, the loss of the measurement is described as a diagonal array type random variable, and the component is a Bernoulli random process and has universality;

(2) based on the observed value of the time-staggered sampling, a sequential fusion algorithm with a serial structure is adopted to design a segmented continuous sequential fusion filter, and a multi-AUV cooperative positioning method with high precision and real-time performance is provided, and the method not only comprises an accurate estimated value at the sampling moment, but also comprises an accurate estimated value in a sampling interval. Compared with centralized fusion filtering, the method effectively reduces the calculation burden, reduces the calculation complexity and is more suitable for actual engineering.

(3) The invention is suitable for a large-scale multi-AUV formation system.

Drawings

FIG. 1 is a schematic structural diagram of a master-slave multi-AUV co-location model according to the present invention;

FIG. 2 is a block diagram of the basic structure of an underwater acoustic communication network according to the present invention;

FIG. 3 is a schematic view of a sequential fusion multi-AUV co-location process based on measurement information staggered time sampling according to the present invention;

FIG. 4 shows the actual state of the multi-AUV co-location system from the x-axis of the AUV and its estimated value in an embodiment of the present invention;

FIG. 5 shows the actual linear velocity of the multi-AUV co-located system from the x-axis of the AUV and its estimated value in the embodiment of the present invention;

FIG. 6 shows the true state of the multi-AUV co-location system from the y-axis of the AUV and its estimated value in an embodiment of the present invention;

FIG. 7 shows the true linear velocity of the multi-AUV co-location system from the Y-axis of the AUV and its estimated value in an embodiment of the present invention;

Detailed Description

The present invention will be described in further detail with reference to the accompanying drawings and examples.

As shown in fig. 1 to 7, a method for co-locating multiple AUVs in an underwater acoustic communication system includes the following steps:

master-slave multi-AUV co-location mode as shown in fig. 1, when a location period starts, a communication request signal is sent from the AUV to the master AUV1, and the master AUV1 broadcasts its own location information to the slave AUV using a hydroacoustic Modem. The main AUV position information is obtained from the AUV reverse decoding and the relative distance information is obtained by ranging techniques. Using the same principle of operation, the position information of the master AUV2 and the relative distance information between them are obtained from the AUV. And the slave AUV utilizes the received information to realize the positioning of the position of the slave AUV based on a corresponding co-positioning filtering algorithm. Due to the complex underwater environment and the special narrowband property, multipath property and fading property of the underwater acoustic communication network, the underwater acoustic signal is attenuated or even interfered when being transmitted in the underwater acoustic communication channel, so that the measurement positioning information of the main AUV is delayed or lost, and the like, and the structure of the underwater acoustic communication network is shown in FIG. 2.

The actual multi-AUV collaborative navigation positioning in the underwater acoustic communication network is reasonably abstracted into a mathematical model according to the principle, and the following conditions are required to be met:

(1) the AUV system in the same horizontal plane is considered, so that the AUV is provided with the depth pressure sensor, the AUV depth can be independently and accurately measured, the distances of UUV with different depths can be projected to the horizontal plane, and the condition that the two AUVs are overlapped on the horizontal plane is not considered. The AUV has good static stability and is in a balanced state;

(2) in order to ensure that the slave AUVs can sequence the received information in sequence according to the time sequence, the master AUV and the slave AUVs need to be subjected to clock synchronization;

(3) the underwater sound Modem has an independent distance measurement mode, so that the distance obtained in the underwater sound distance measurement process is assumed to be relatively accurate;

(4) aiming at the narrowband characteristic of underwater acoustic communication, in order to avoid network congestion caused by channel sharing, a time division multiplexing mechanism is adopted to share a channel in an MAC layer, different time slots are allocated to different AUVs, and a time-staggered sampling and time-staggered communication mode is adopted. According to the characteristics of master-slave type cooperative positioning, the slave AUV respectively receives the measurement information from different master AUVs, and the measurement information of the master AUV is sequentially utilized to complete self positioning according to the information sending time sequence. The present invention ignores the situation of measurement information delay.

The multi-AUV co-location process in the underwater acoustic communication network can be characterized as a linear sampling system, namely, the mathematical model of the underwater multi-AUV co-location system is as follows:

wherein X (t) e Rⁿ，U(t)∈R^mRespectively representing the position information vector and the control input vector of an AUV in the underwater acoustic communication system, wherein t is more than or equal to 0;indicates the l th received from AUV_iMeasurement information of individual main AUV, where k_iDenotes the l_iThe time when the main AUV measurement information arrives is k for every two AUV sampling intervals_i+1-k_iFor convenience, h is assumed to be 1; w (t) ε R^pRepresenting system process noise, including velocity measurement noise and orientation measurement noise;representing underwater acoustic distance measurement noise between the AUVs; diagonal matrix type random variableDescribing a metrology loss profile, wherein the componentsObeying a Bernoulli distribution with a probability distribution ofi，j＝1，…m。 Meaning that the metrology data was successfully received,meaning that no metrology data is received but a metrology noise term is present in the observation. Specifically, the measurement information is transmitted by time-staggered sampling and time-staggered communication, so that the first slave AUV has one sampling period (k, k + 1)]Sequentially receiving the measurement information of m main AUVs, and numbering l according to the sequence₁，...，l_mThe ideal discrete time of the measurement arriving in sequence is marked as k₁，k₂，...，k_mThen the relationship is satisfied: k is less than k₁＜k₂＜…＜k_mK + 1. A, B, G, H are respectively appropriate dimension constant matrixes. w (t) is wiener with zero mean and variance matrix of QA process;is a zero mean and square difference matrix ofWhite gaussian noise of (1); the initial value X (0) of the state of the system is that the mean value is zero and the variance is V₀And X (0) and w (t),andl_i＝l₁，...，l_mare independent of each other.

Based on the mathematical model, the method discloses a multi-AUV cooperative positioning method of an underwater acoustic communication system based on a sequential fusion algorithm, which comprises the following steps:

(1) the mathematical model is used as a mathematical model of the underwater multi-AUV cooperative positioning system;

(2) determining a sampling period (k, k + 1)]The first slave AUV receives the measurement information of m master AUVs in turn in a sampling period, that isk＜k₁＜k₂＜…＜k_m＝k+1；

(3) Based on Kalman filtering algorithm, single measurement information is givenThe local filter of (1);

(4) and (3) adopting a sequential fusion method, gradually processing observation data in a sampling period according to the sequence of local measurement information reaching a fusion estimation center, giving a fusion estimation value in real time, and finally giving an accurate estimation value of the pose information of the AUV at the current moment, as shown in figure 3.

The specific implementation method comprises the following steps:

step 1: the given time t is more than or equal to 0, and for the sampling system (1), the system can be rewritten into a system with the effect of additive disturbance and random loss observation in the sampling period (k, k + 1)

Wherein the content of the first and second substances,satisfy the requirement of

Here, the first and second liquid crystal display panels are,

step 2: based on the first arriving₁Individual main AUV measurement informationUsing kalman filtering algorithm, t ═ k can be obtained₁Temporal local filterSum filter error variance matrix

Wherein the initial value isA filter gain array is represented by a matrix of filter gains,and (3) a filtering error variance matrix is represented, and the following equation is satisfied:

here, the initial value isS(k₁) Satisfies the following Lyapunov equation:

the initial value is S (k)₀)＝S(k)，k₀＜t≤k₁。

And step 3: for k₁＜t≤k₂Based on the l₂Individual main AUV measurement informationUsing kalman filtering algorithm, t ═ k can be obtained₂Temporal local filterSum filter error variance matrixThe local predictor obtains t as k₂A predicted value of the time;

wherein the initial value is t ═ k₁Filtered value of time of dayA filter gain array is represented by a matrix of filter gains,and (3) representing a filter error square difference matrix, and satisfying the following equation:

here, the initial value is

And 4, step 4: and so on, based on the measurement information of each main AUV arriving in sequenceCan obtain the first_mA filterWhen t is k +1, the l-th state-optimal sequential fusion filter from the AUV is obtained as

And 5: and k is k +1, and the step 1 is returned.

Due to the fact that the measurement information adopts a transmission mode of time-staggered sampling and time-staggered communication, the measurement information is asynchronous, the traditional centralized fusion mode is difficult to be applied to practical application, measurement and dimension expansion bring large calculation burden, measurement noise after dimension expansion is correlated with process noise, and complexity of algorithm design is increased. At the moment, a sequential fusion algorithm is adopted, observation data are processed step by step according to the sequence of local observation data reaching a fusion center, a fusion estimation value is given in real time, the algorithm is executed without waiting for all data to reach the fusion center, the calculation burden can be effectively reduced, and the calculation complexity is reduced.

The invention is further illustrated below with reference to specific examples. In the experiment, the situation that measurement is lost in the underwater acoustic communication network is considered, and two main AUVs are adopted to alternately perform cooperative positioning on the auxiliary AUVs. Specifically, a set of parameters is obtained from the actual engineering as follows:

the invention adoptsThe sampling time is set to T-30 s. Initial value of state is

Further, with the proposed sequential fusion algorithm based multi-AUV co-location method, the position of the AUV, i.e. the x-axis state value and linear velocity, and the y-axis state value and linear velocity, can be obtained, as shown in fig. 4-7.

Firstly, establishing a continuous state equation of a multi-AUV motion process and a discrete measurement equation with measurement loss in an underwater acoustic communication network to obtain system parameters; then, obtaining an optimal local filter based on a Kalman filtering algorithm, wherein the filter is continuous in a segmented mode and comprises an estimated value at a sampling moment and an accurate estimated value of a sampling interval; and finally, designing a segmented continuous sequential fusion filter based on a sequential fusion algorithm, and providing a multi-AUV co-location method with high precision and real-time performance. The invention designs the multi-AUV cooperative positioning method of the underwater acoustic communication system with loss measurement by utilizing the Kalman filtering algorithm and the sequential fusion method, thereby not only reducing the calculation burden, but also improving the real-time performance of positioning.

The above description only describes the embodiments of the present invention, but not limits the scope of the present invention, and those skilled in the art should understand that various modifications or variations can be made without inventive effort by those skilled in the art based on the technical solution of the present invention.