A multi-robot collaborative positioning method, processor and medium for communication interference

Abstract

The present invention relates to the technical field of multi-robot collaborative positioning, and discloses a multi-robot collaborative positioning method, processor and medium facing communication interference. Aiming at problems such as communication interruption and communication packet loss during collaborative positioning, a flexible collaborative positioning algorithm is designed, that is, a distributed collaborative positioning method combining global and local. The method divides the state update process of the robot during collaborative positioning into three parts: time propagation update, observation update and communication update. The time state update is updated by global state estimation, and a time prediction update model under the communication interference state is established; the observation update is completed by combining the local and global, and the event trigger condition is constructed by using the error relationship, and the event trigger mechanism is constructed by the condition to make a decision on the observed object; the local state update is combined with the CI algorithm to complete the communication update; the consistency, robustness and communication cost of the system are balanced, and a hybrid collaborative positioning method coupled with multiple estimation methods is constructed.

Description

Technical Field

[0001] The present invention relates to the technical field of multi-robot collaborative positioning, and in particular to a multi-robot collaborative positioning method, processor and medium facing communication interference. Background Art

[0002] Positioning is the key to the autonomous completion of tasks by mobile robots and the prerequisite for robot mapping and navigation. When absolute positioning information cannot be obtained, the robot can only rely on its own sensors to provide positioning information in the form of dead reckoning. However, similar positioning methods are susceptible to error accumulation, and ultimately obtain positioning information with huge errors or even errors. Multi-robot systems have strong scalability. The collaborative positioning algorithm constructed by relative observation and information sharing between robots can effectively suppress the cumulative growth of errors. However, international research on collaborative positioning is currently in the algorithm design stage, and some focus on improving theoretical proofs, while there are still some deficiencies in the exploration of actual engineering problems. In particular, when there is a communication failure or even a strong communication failure in the system, the traditional collaborative positioning scheme generally does not perform well.

[0003] Therefore, a method with strong robustness and high positioning accuracy in the face of communication interference is needed to realize the positioning solution of multi-robot system. Summary of the invention

[0004] 1. Technical issues to be resolved

[0005] In view of the shortcomings of the prior art, the present invention provides a multi-robot collaborative positioning method, processor and medium for communication interference, which have the advantages of balancing the consistency, robustness and communication cost of the system, and solve the above-mentioned technical problems.

[0006] (II) Technical solution

[0007] To achieve the above object, the present invention provides the following technical solution: a multi-robot collaborative positioning method for communication interference, comprising the following steps:

[0008] S1. Split the robot's state update process during collaborative positioning into three processes, namely time propagation update, observation update and communication update;

[0009] S2. Estimate the status of all robots. For robots that are unable to obtain positioning information due to communication interference, establish a time prediction update model to estimate their status. Specifically, the following steps are included:

[0010] S2.1. For the observed The global state of each robot is represented by the following expression:

[0011]

[0012] in, express Global status information at all times, , , , Respectively represent robots , , , exist The location information obtained at all times can be used to prevent communication interference. , , , Represents a robot , , , exist The predicted position at the time, , Represents a robot exist The predicted position at the time, superscript represents transpose, Represents a robot exist The transposition of the state at time , and Represents a robot exist The position coordinates at the time, express exist The heading angle at the moment;

[0013] S2.2. For robots exist Status of the moment Perform time propagation update and obtain the global state time update. The specific expression is as follows:

[0014]

[0015] in, Represents a robot exist The state obtained by the time propagation update is Represents a robot exist Time speed input, Represents a robot exist The moment speed input noise, Indicates unit time, express The cosine value of express The sine value of Represents a robot Time propagation update equation, represents the lateral position information updated by time propagation, represents the longitudinal position information updated by time propagation, The heading information obtained by time propagation update and the expression of global state time update are as follows:

[0016]

[0017] in, The time propagation update equation representing the position and input velocity of robot 1 is, The time propagation update equation representing the position and input velocity of robot j is, The time-propagation update equation representing the position and input velocity of robot N is, Represents a robot exist The predicted speed at the moment, Represents the time propagation update equation of robot position and input velocity, Indicates global state time update;

[0018] S2.3. Being a robot When communication is disturbed, the robot exist Prediction speed at the moment Replaced by the covariance matrix Extremely large Gaussian random vector, when the robot When communication is restored, update the robot The updated expression of the covariance matrix during time propagation update is as follows:

[0019]

[0020] in, Represents a robot exist The covariance matrix at the time of time propagation update, Represents a robot exist The covariance matrix at the time of time propagation update, represents the state transition matrix under global state update, is the process noise, and the covariance matrix of the global state update is obtained by solving it. The expression is as follows:

[0021]

[0022] in, Represents a robot exist The covariance matrix of the local state at time , Represents a robot exist The covariance matrix of the local state at time , Represents a robot The state transition matrix, Represents a robot The transpose of the state transfer matrix, represents the global process noise, Represents the noise transfer matrix of robot i, and solves to obtain the covariance matrix of local state update:

[0023]

[0024]

[0025] in, The diagonal elements are The diagonal matrix of Respectively represent robots Different state transition matrices, , , denote the noise transfer matrices of robots 1, i, and N respectively, with the superscript represents transpose, and , , , The expression is as follows:

[0026]

[0027]

[0028]

[0029]

[0030] in, represents the time propagation update equation, Indicates status information. Indicates the input speed, Location information;

[0031] S3. Different estimation methods are established according to whether the robot's observation targets are robots or fixed landmarks. When the observation target is a robot, the global state EKF algorithm is used to estimate the pose. When the observation target is a landmark, the local state EKF algorithm is used to update the positioning information. When the observation targets are both robots and fixed landmarks, the event triggering condition is constructed by estimating the error, and then the event triggering observation update is established. The specific steps are as follows:

[0032] S3.1. Construct the observation equation. The specific expression is as follows:

[0033] The observation target is the observation equation of the robot:

[0034]

[0035]

[0036] in, Represents the observation equation when the observed object is a robot, represents the predicted position information of robot j, Represents a robot exist The state obtained by the time propagation update is and denote the predicted lateral and longitudinal positions of robot j, respectively, and They represent the time propagation update of robot i in The horizontal and vertical position information at the moment, Indicates that robot i time propagation update is Time heading information, represents the relative position relationship between robot i and robot j calculated by time propagation update, Represents a robot and robots exist The estimated relative angle at the moment, Represents a robot and robots exist The estimated relative distance at the moment, the relative distance and relative angle between the robots are obtained through sensors:

[0037]

[0038] in, The sensor observes robot i and robot j The relative position of time, Represents a robot and robots exist The relative distance of time, Represents a robot and robots exist The relative angle of the moment;

[0039] The observation target is a landmark The observation equation is:

[0040]

[0041]

[0042] in, Indicates that the observed object is a landmark The observation equation when Indicates the location of a landmark. Represents robot i and landmark Time propagation updates the estimated relative position relationship, Represents a robot and landmarks exist The estimated relative angle at the moment, Represents a robot and landmarks exist The estimated relative distance between the robot and the landmark at the moment is obtained by the sensor The measured value expression is as follows:

[0043]

[0044] in, Indicates the distance between robot i and landmark measured by the sensor The relative position relationship between Represents a robot With landmarks Between The relative distance of time, Represents a robot With landmarks Between The relative angle of the moment;

[0045] S3.2. Construct an event-triggered observation update strategy: When the robot and the landmark are both observed, a binary decision variable is constructed to determine the estimated credibility of the landmark observation update. The specific expression is as follows:

[0046]

[0047] in, represents a binary decision variable, Represents a robot With landmarks Between The relative distance of time, Indicated in Estimated relative distance at each moment and the measured relative distance The error value between " indicates updating using landmark observations," " indicates the use of robot observation updates, and They represent time trigger parameters respectively, and the event trigger observation update expression is as follows:

[0048]

[0049] in, express Global status information at all times, Indicates time update status information, express The gain coefficient at time, express The moment is the relative position relationship observed by the sensor. express Moment is the time propagation update to estimate the relative position of the observation, and , , , The expression is as follows:

[0050]

[0051]

[0052]

[0053]

[0054] in, represents the global state information updated by time propagation, represents the relative position relationship between robot i and robot j calculated by time propagation update, Representation of robot i and landmark The relative position relationship calculated by the time propagation update is represents the relative position relationship between robot i and robot j observed by the actual sensor, represents the relative position relationship between robot i and landmark 𝜆 measured by the sensor, represents the gain coefficient when the observed object is robot j, Represents the gain coefficient when the observed object is landmark 𝜆, and The expression is as follows:

[0055]

[0056]

[0057] in, Represents a robot exist The covariance matrix at the time of time propagation update, Represents a robot exist The covariance matrix of the local state at time , Represents state variables to the observed object as a robot The transformation matrix in the global state at time , Representing state variables to observation objects as landmarks The transformation matrix in the global state at time , express The transpose of express The transpose of represents the inverse of the matrix, express The noise matrix is ​​measured at all times and the covariance matrix is ​​updated. The specific expression is as follows:

[0058]

[0059]

[0060] in, express The inverse matrix of the moment measurement noise matrix, express The inverse matrix of the moment measurement noise matrix;

[0061] S4, the communication update is specifically to use local state estimation combined with CI algorithm to perform communication update, including the following steps:

[0062] S4.1. According to the calculation formula, the observed robot The position of the observation robot The location information is as follows:

[0063]

[0064] in, Represents the observation robot By the observed robot Estimated location information, Indicates the relative distance and relative angle measured by the sensor. Indicates that robot j is Location information at the moment, Denotes the communication update equation, the observation robot By the observed robot The specific calculation expression of the estimated position information is as follows:

[0065]

[0066] in, Represents a robot and robots exist The relative distance of time, represents the relative angle between robot i and robot j, Indicates communication update robot and robots The inferred lateral position information, Indicates communication update robot and robots Infer the axial position information;

[0067] S4.2. When multiple robots and robots When communicating, robots The state update expression is as follows:

[0068]

[0069]

[0070] in, Representation and Robot Communication robots set, represents the fitting coefficient and , superscript represents the inverse of the matrix, express The global state covariance matrix at time t, Representation and Robot The communicating robots collect all the data and sum them up. , and meets the minimization performance index, the expression is as follows:

[0071]

[0072] in, represents the minimum value, Indicates that the subscript fitting coefficients are obeyed The minimum value required, represents the trace of the interior matrix.

[0073] As a preferred technical solution of the present invention, the step S3.2 and The expression is as follows:

[0074]

[0075]

[0076] in, represents the matrix, represents the observation equation when the observed object is landmark λ, Indicates the location of a landmark. and Respectively The local state transfer matrix of robot i and robot j is expressed as follows:

[0077]

[0078]

[0079] in, Represents the observation equation when the observed object is a robot, Indicates that robot i is The state obtained by the time propagation update is Indicates that robot j is The updated position information is propagated at the moment.

[0080] The present invention also provides a processor, which is used to run a program, wherein the program executes the above-mentioned multi-robot collaborative positioning method oriented to communication interference when running.

[0081] The present invention also provides a computer-readable storage medium, which includes a stored program, wherein when the program is running, the device where the computer-readable storage medium is located is controlled to execute the above-mentioned multi-robot collaborative positioning method oriented to communication interference.

[0082] Compared with the prior art, the present invention provides a multi-robot collaborative positioning method, processor and medium for communication interference, which has the following beneficial effects:

[0083] The algorithm proposed in the present invention has excellent or relatively good comprehensive performance balance in most communication situations. Compared with traditional algorithms based on local state estimation and centralized algorithms, the algorithm of the present invention has significant ability to suppress communication interference, and shows relatively ideal effects under the interference of communication interruption and communication packet loss. It has similar effects to the algorithm based on global state estimation in terms of communication elasticity, but the algorithm of the present invention has obvious advantages in communication cost, estimation performance when communication is normal, and estimation uncertainty. LS-CI and LS-Cen only show relatively ideal performance in individual aspects, and their performance in resisting communication interference is poor, especially LS-Cen, which is highly dependent on communication, is most affected. BRIEF DESCRIPTION OF THE DRAWINGS

[0084] Figure 1 is a flow chart of the method of the present invention;

[0085] Figure 2 It is a schematic diagram of the influence of the event triggering parameters on the starting point of the present invention;

[0086] Figure 3 This is a schematic diagram of the present invention's anti-communication interference;

[0087] Figure 4 2. It is a schematic diagram for comparing the anti-communication packet loss rate of the present invention;

[0088] Figure 5 It is a schematic diagram of the communication cost comparison of the present invention;

[0089] Figure 6 It is a schematic diagram of the estimated uncertainty comparison of the present invention. DETAILED DESCRIPTION

[0090] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0091] See also Figure 1-6 , a multi-robot collaborative positioning method for communication interference, comprising the following steps:

[0092] S1. Split the robot's state update process during collaborative positioning into three processes, namely time propagation update, observation update and communication update;

[0093] S2. Estimate the status of all robots. For robots that are unable to obtain positioning information due to communication interference, establish a time prediction update model to estimate their status. Specifically, the following steps are included:

[0094] S2.1. For the observed The global state of each robot is represented by the following expression:

[0095]

[0096] in, express Global status information at all times, , , , Respectively represent robots , , , exist The location information obtained at all times can be used to prevent communication interference. , , , Represents a robot , , , exist The predicted position at the time, , Represents a robot exist The predicted position at the time, superscript represents transpose, Represents a robot exist The transposition of the state at time , and Represents a robot exist The position coordinates at the time, express exist The heading angle at the moment;

[0097] S2.2. For robots exist Status of the moment Perform time propagation update and obtain the global state time update. The specific expression is as follows:

[0098]

[0099] in, Represents a robot exist The state obtained by the time propagation update is Represents a robot exist Time speed input, Represents a robot exist The moment speed input noise, Indicates unit time, express The cosine value of express The sine value of Represents a robot Time propagation update equation, represents the lateral position information updated by time propagation, represents the longitudinal position information updated by time propagation, The heading information obtained by time propagation update and the expression of global state time update are as follows:

[0100]

[0101] in, The time propagation update equation representing the position and input velocity of robot 1 is, The time propagation update equation representing the position and input velocity of robot j is, The time-propagation update equation representing the position and input velocity of robot N is, Represents a robot exist The predicted speed at the moment, The time propagation update equation representing the position and input velocity of robot i is, Indicates global state time update;

[0102] S2.3. Being a robot When communication is disturbed, the robot exist Prediction speed at the moment Replaced by the covariance matrix Extremely large Gaussian random vector, when the robot When communication is restored, update the robot The updated expression of the covariance matrix during time propagation update is as follows:

[0103]

[0104] in, Represents a robot exist The covariance matrix at the time of time propagation update, Represents a robot exist The covariance matrix at the time of time propagation update, represents the state transition matrix under global state update, is the process noise, and the covariance matrix of the global state update is obtained by solving it. The expression is as follows:

[0105]

[0106] in, Represents a robot exist The covariance matrix of the local state at time , Represents a robot exist The covariance matrix of the local state at time , Represents a robot The state transition matrix, Represents a robot The transpose of the state transfer matrix, represents the global process noise, Represents the noise transfer matrix of robot i, and solves to obtain the covariance matrix of local state update:

[0107]

[0108]

[0109] in, The diagonal elements are The diagonal matrix of Respectively represent robots Different state transition matrices, , , denote the noise transfer matrices of robots 1, i, and N respectively, and , , , The expression is as follows:

[0110]

[0111]

[0112]

[0113]

[0114] in, represents the time propagation update equation, Indicates status information. Indicates the input speed, Location information;

[0115] S3. Different estimation methods are established according to whether the robot's observation targets are robots or fixed landmarks. When the observation target is a robot, the global state EKF algorithm is used to estimate the pose. When the observation target is a landmark, the local state EKF algorithm is used to update the positioning information. When the observation targets are both robots and fixed landmarks, the event triggering condition is constructed by estimating the error, and then the event triggering observation update is established. The specific steps are as follows:

[0116] S3.1. Construct the observation equation. The specific expression is as follows:

[0117] The observation target is the observation equation of the robot:

[0118]

[0119]

[0120] in, Represents the observation equation when the observed object is a robot, represents the predicted position information of robot j, Represents a robot exist The state obtained by the time propagation update is and denote the predicted lateral and longitudinal positions of robot j, respectively, and They represent the time propagation update of robot i in The horizontal and vertical position information at the moment, Indicates that robot i time propagation update is Time heading information, represents the relative position relationship between robot i and robot j calculated by time propagation update, Represents a robot and robots exist The estimated relative angle at the moment, Represents a robot and robots exist The estimated relative distance at the moment, the relative distance and relative angle between the robots are obtained through sensors:

[0121]

[0122] in, The sensor observes robot i and robot j The relative position of time, Represents a robot and robots exist The relative distance of time, Represents a robot and robots exist The relative angle of the moment;

[0123] The observation target is a landmark The observation equation is:

[0124]

[0125]

[0126] in, Indicates that the observed object is a landmark The observation equation when Indicates the location of a landmark. Represents robot i and landmark Time propagation updates the estimated relative position relationship, Represents a robot and landmarks exist The estimated relative angle at the moment, Represents a robot and landmarks exist The estimated relative distance between the robot and the landmark at the moment is obtained by the sensor The measured value expression is as follows:

[0127]

[0128] in, Indicates the distance between robot i and landmark measured by the sensor The relative position relationship between Represents a robot With landmarks Between The relative distance of time, Represents a robot With landmarks Between The relative angle of the moment;

[0129] S3.2. Construct an event-triggered observation update strategy: When the robot and the landmark are both observed, a binary decision variable is constructed to determine the estimated credibility of the landmark observation update. The specific expression is as follows:

[0130]

[0131] in, represents a binary decision variable, Represents a robot With landmarks Between The relative distance of time, Indicated in Estimated relative distance at each moment and the measured relative distance The error value between " indicates updating using landmark observations," " indicates the use of robot observation updates, and Respectively represent the time trigger parameters, , event trigger parameters , The smaller the selection, the easier it is to trigger the event. = 0, you can get a trigger parameter only The absolute trigger conditions are designed = 0, you can get a The relative trigger condition. The larger the value, the lower the credibility of using landmarks to suppress errors. We should avoid using landmarks to update the current state. The expression for event-triggered observation update is as follows:

[0132]

[0133] in, express Global status information at all times, Indicates time update status information, express The gain coefficient at time, express The moment is the relative position relationship observed by the sensor. express Moment is the time propagation update to estimate the relative position of the observation, and , , , The expression is as follows:

[0134]

[0135]

[0136]

[0137]

[0138] in, represents the global state information updated by time propagation, represents the relative position relationship between robot i and robot j calculated by time propagation update, Representation of robot i and landmark The relative position relationship calculated by the time propagation update is represents the relative position relationship between robot i and robot j observed by the actual sensor, represents the relative position relationship between robot i and landmark 𝜆 measured by the sensor, represents the gain coefficient when the observed object is robot j, represents the gain coefficient when the observed object is landmark 𝜆, and The expression is as follows:

[0139]

[0140]

[0141] in, Represents a robot exist The covariance matrix at the time of time propagation update, Represents a robot exist The covariance matrix of the local state at time , Represents state variables to the observed object as a robot The transformation matrix in the global state at time , Representing state variables to observation objects as landmarks The transformation matrix in the global state at time , express The transpose of express The transpose of represents the inverse of the matrix, express The noise matrix is ​​measured at all times and the covariance matrix is ​​updated. The specific expression is as follows:

[0142]

[0143]

[0144] in, express The inverse matrix of the moment measurement noise matrix, express The inverse matrix of the moment measurement noise matrix and The expression is as follows:

[0145]

[0146]

[0147] in, represents the matrix, represents the observation equation when the observed object is landmark λ, Indicates the location of a landmark. and Respectively The local state transfer matrix of robot i and robot j is expressed as follows:

[0148]

[0149]

[0150] in, Represents the observation equation when the observed object is a robot, Indicates that robot i is The state obtained by the time propagation update is Indicates that robot j is The position information obtained by the momentary time propagation update;

[0151] S4. Considering that the robot communication update is related to the observation information, the constraint of the observation information on the communication update is retained. In fact, communication with the target is meaningful only when the information between the target and the robot is observed. Therefore, communication is only performed on robots that meet the communication conditions and obtain measurement information, which can reduce the communication pressure. To ensure the consistency of the estimation, the communication update is specifically performed by using local state estimation combined with the CI algorithm, including the following steps:

[0152] S4.1. According to the calculation formula, the observed robot The position of the observation robot The location information is as follows:

[0153]

[0154] in, Represents the observation robot By the observed robot Estimated location information, Indicates the relative distance and relative angle measured by the sensor. Indicates that robot j is Location information at the moment, represents the communication update equation, observing the robot By the observed robot The specific calculation expression of the estimated position information is as follows:

[0155]

[0156] in, Represents a robot and robots exist The relative distance of time, represents the relative angle between robot i and robot j, Indicates communication update robot and robots The inferred lateral position information, Indicates communication update robot and robots Infer the axial position information;

[0157] S4.2. When multiple robots and robots When communicating, robots The state update expression is as follows:

[0158]

[0159]

[0160] in, Representation and Robot Communication robots set, represents the fitting coefficient and , superscript represents the inverse of the matrix, express The global state covariance matrix at time t, Representation and Robot The communicating robots collect all the data and sum them up. , and meets the minimization performance index, the expression is as follows:

[0161]

[0162] in, represents the minimum value, Indicates that the subscript fitting coefficients are obeyed The minimum value required, represents the trace of the interior matrix.

[0163] At this point, the communication update part has been completed. The purpose of the communication update is to further limit the error growth of the positioning information obtained from the observation update. Its ability to suppress errors is significant. By combining global state updates with local state updates, the dependence of observation updates on communication information is decoupled, and finally the collaborative positioning requirements under communication interference are achieved. In the present invention, only the collaborative positioning from time t to time t+1 is elaborated in detail. The steps of extending time t+1 to time t+2 are similar and will not be elaborated in detail here. However, it is a cyclic process. Through multiple cycles within a time interval, continuous collaborative positioning within a period of time is finally achieved, which still belongs to the protection scope of the present invention. The present invention Figure 2 The figure shows the impact of different event trigger parameters on the starting point of the five robots R1~R5.

[0164] The global-local-state covariance intersection (GLS-CI) positioning algorithm proposed in this paper is verified and analyzed based on the dataset published by the University of Toronto. The dataset has 9 sub-datasets. The verification time is 720s, and the time accuracy of the verification is 1s. The number of robots in the whole system is set to 6. During the verification, the following three publicly reported algorithms will be compared and analyzed:

[0165] (1) Local State Estimation Distributed Co-localization Algorithm (LS-CI);

[0166] (2) Global State Estimation Distributed Co-localization Algorithm (GS-CI);

[0167] (3) Local state estimation centralized collaborative localization algorithm (LS-Cen).

[0168] The obtained partial positioning performance and anti-communication capability are shown in the attached figure. Figure 3 ~ Figure 6 This part is mainly completed using data set 6; under normal communication, the average root mean square error (ARMSE, unit: m) of the 9 sub-data sets is shown in the following table.

[0169]

[0170] The present invention also provides a processor, which is used to run a program, wherein the program executes the above-mentioned multi-robot collaborative positioning method oriented to communication interference when running.

[0171] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium includes a stored program, wherein when the program is running, the device where the computer-readable storage medium is located is controlled to execute the above-mentioned multi-robot collaborative positioning method facing communication interference

[0172] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A multi-robot collaborative positioning method for communication interference, characterized by: The following steps are involved: S1. Split the robot's state update process during collaborative positioning into three processes, namely time propagation update, observation update and communication update; S2. Estimate the status of all robots. For robots that are unable to obtain positioning information due to communication interference, establish a time prediction update model to estimate their status. Specifically, the following steps are included: S2.

1. For the observed The global state of each robot is represented by the following expression: ； in, express Global status information at all times, , , , Respectively represent robots , , , exist The location information obtained at all times can be used to prevent communication interference. , , , Represents a robot , , , exist The predicted position at the time, , Represents a robot exist The predicted position at the time, superscript represents transpose, Represents a robot exist The transposition of the state at time , and Represents a robot exist The position coordinates at the time, express exist The heading angle at the moment; S2.

2. For robots exist Status of the moment Perform time propagation update and obtain the global state time update. The specific expression is as follows: ； in, Represents a robot exist The state obtained by the time propagation update is Represents a robot exist Time speed input, Represents a robot exist The moment speed input noise, Indicates the unit time, express The cosine value of express The sine value of Represents a robot Time propagation update equation, represents the lateral position information updated by time propagation, represents the longitudinal position information updated by time propagation, The heading information obtained by time propagation update and the expression of global state time update are as follows: ； in, The time propagation update equation representing the position and input velocity of robot 1 is, The time propagation update equation representing the position and input velocity of robot j is, The time-propagation update equation representing the position and input velocity of robot N is, Represents a robot exist The predicted speed at the moment, The time propagation update equation representing the position and input velocity of robot i is, Indicates the global state time update; S2.

3. Being a robot When communication is disturbed, the robot exist Prediction speed at the moment Replaced by the covariance matrix Extremely large Gaussian random vector, when the robot When communication is restored, update the robot The updated expression of the covariance matrix during time propagation update is as follows: ； in, Represents a robot exist The covariance matrix at the time of time propagation update, Represents a robot exist The covariance matrix at the time of time propagation update, represents the state transition matrix under global state update, is process noise, and the covariance matrix of the global state update can be obtained according to the following formula: ； in, Represents a robot exist The covariance matrix of the local state at time , Represents a robot exist The covariance matrix of the local state at time , Represents a robot The state transition matrix, Represents a robot The transpose of the state transfer matrix, represents the global process noise, Represents the noise transfer matrix of robot i, and solves to obtain the covariance matrix of local state update: ； ； in, The diagonal elements are The diagonal matrix of Respectively represent robots Different state transition matrices, , , denote the noise transfer matrices of robots 1, i, and N respectively, and , , , The expression is as follows: ； ； ； ； in, represents the time propagation update equation, Indicates status information. Indicates the input speed, Location information; S3. Different estimation methods are established according to whether the robot's observation targets are robots or fixed landmarks. When the observation target is a robot, the global state EKF algorithm is used to estimate the pose. When the observation target is a landmark, the local state EKF algorithm is used to update the positioning information. When the observation targets are both robots and fixed landmarks, the event triggering condition is constructed by estimating the error, and then the event triggering observation update is established. The specific steps are as follows: S3.

1. Construct the observation equation. The specific expression is as follows: The observation target is the observation equation of the robot: ； ； in, Represents the observation equation when the observed object is a robot, represents the predicted position information of robot j, Represents a robot exist The state obtained by the time propagation update is and denote the predicted lateral and longitudinal positions of robot j, respectively, and They represent the time propagation update of robot i in The horizontal and vertical position information at the moment, Indicates that robot i time propagation update is Time heading information, represents the relative position relationship between robot i and robot j calculated by time propagation update, Represents a robot and robots exist The estimated relative angle at the moment, Represents a robot and robots exist The estimated relative distance at the moment, the relative distance and relative angle between the robots are obtained through sensors: ； in, The sensor observes robot i and robot j The relative position of time, Represents a robot and robots exist The relative distance of time, Represents a robot and robots exist The relative angle of the moment; The observation target is a landmark The observation equation is: ； ； in, Indicates that the observed object is a landmark The observation equation when Indicates the location of a landmark. Represents robot i and landmark Time propagation updates the estimated relative position relationship, Represents a robot and landmarks exist The estimated relative angle at the moment, Represents a robot and landmarks exist The estimated relative distance between the robot and the landmark at the moment is obtained by the sensor The measured value expression is as follows: ； in, Indicates the distance between robot i and landmark measured by the sensor The relative position relationship between Represents a robot With landmarks Between The relative distance of time, Represents a robot With landmarks Between The relative angle of the moment; S3.

2. Construct an event-triggered observation update strategy: When the robot and the landmark are both observed, a binary decision variable is constructed to determine the estimated credibility of the landmark observation update. The specific expression is as follows: ； in, represents a binary decision variable, Represents a robot With landmarks Between The relative distance of time, Indicated in Estimated relative distance at each moment and the measured relative distance The error value between " indicates updating using landmark observations, " " indicates the use of robot observation updates, and They represent time trigger parameters respectively, and the event trigger observation update expression is as follows: ； in, express Global status information at all times, Indicates time update status information, express The gain coefficient at time, express The moment is the relative position relationship observed by the sensor. express Moment is the time propagation update to estimate the relative position of the observation, and , , , The expression is as follows: ； ； ； ； in, represents the global state information updated by time propagation, represents the relative position relationship between robot i and robot j calculated by time propagation update, Representation of robot i and landmark The relative position relationship calculated by the time propagation update is represents the relative position relationship between robot i and robot j observed by the actual sensor, represents the relative position relationship between robot i and landmark 𝜆 measured by the sensor, represents the gain coefficient when the observed object is robot j, Represents the gain coefficient when the observed object is landmark 𝜆, and The expression is as follows: ； ； in, Represents a robot exist The covariance matrix at the time of time propagation update, Represents a robot exist The covariance matrix of the local state at time , Represents state variables to the observed object as a robot The transformation matrix in the global state at time , Representing state variables to observation objects as landmarks The transformation matrix in the global state at time , express The transpose of express The transpose of represents the inverse of the matrix, express The noise matrix is ​​measured at all times and the covariance matrix is ​​updated. The specific expression is as follows: ； ； in, express The inverse matrix of the moment measurement noise matrix, express The inverse matrix of the moment measurement noise matrix; S4, the communication update is specifically to use local state estimation combined with CI algorithm to perform communication update, including the following steps: S4.

1. According to the calculation formula, the observed robot The position of the observation robot The location information is as follows: ； in, Represents the observation robot By the observed robot Estimated location information, Indicates the relative distance and relative angle measured by the sensor. Indicates that robot j is Location information at the moment, Denotes the communication update equation, the observation robot By the observed robot The specific calculation expression of the estimated position information is as follows: ； in, Represents a robot and robots exist The relative distance of time, represents the relative angle between robot i and robot j, Indicates communication update robot and robots The inferred lateral position information, Indicates communication update robot and robots Infer the axial position information; S4.

2. When multiple robots and robots When communicating, robots The state update expression is as follows: ； ； in, Representation and Robot Communication robots set, represents the fitting coefficient and , superscript represents the inverse of the matrix, express The global state covariance matrix at time t, Representation and Robot The communicating robots collect all the data and sum them up. , and meets the minimization performance index, the expression is as follows: ； in, represents the minimum value, Indicates that the subscript fitting coefficients are obeyed The minimum value required, represents the trace of the interior matrix.

2. The multi-robot collaborative positioning method facing communication interference according to claim 1, characterized in that: The step S3.2 and The expression is as follows: ； ； in, represents the matrix, represents the observation equation when the observed object is landmark λ, Indicates the location of a landmark. and Respectively The local state transfer matrix of robot i and robot j is expressed as follows: ； ； in, Represents the observation equation when the observed object is a robot, Indicates that robot i is The state obtained by the time propagation update is Indicates that robot j is The updated position information is propagated at the moment.

3. A processor, characterized in that: The processor is used to run a program, wherein when the program is run, a multi-robot collaborative positioning method resistant to communication interference as described in any one of claims 1-2 is executed.

4. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is run, the device where the computer-readable storage medium is located is controlled to execute the multi-robot collaborative positioning method resistant to communication interference as described in any one of claims 1-2.

Images