A collaborative isosurface tracking method for multiple robots with limited communication
By using the field strength and relative position information measured by its own sensor in three-dimensional space for principal component analysis, kinematic controller is designed, and the isosurface tracking problem of multiple robots under communication restricted conditions is solved, realizing the inefficient environment efficient collaborative control and three-dimensional isosurface tracking.
Patent Information
- Application Number
- CN202211156859.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-09-21
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2042-09-21
AI Technical Summary
The existing multi-robot isosurface tracking method cannot achieve collaborative control and efficient decision-making under the conditions of communication limitation, and cannot effectively track isosurfaces in three-dimensional space, especially in extreme environments, which cannot obtain GPS location information and other robots. Sensor information.
By establishing a dynamic model in three-dimensional space, measuring field strength and relative position information with its own sensors, performing principal component analysis to determine the direction of coordinated motion, and designing a kinematic controller to achieve synergistic equality of multiple robots without communication track.
Multi-robot isosurface tracking in a communication-constrained environment, which improves the robustness and flexibility of the system, can adaptively optimize sub-objectives, and enhances the applicability and accuracy of the algorithm by controlling the distance relative to the relative position.
Smart Images

Figure CN115469666B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of multi-robot collaborative environment monitoring, and mainly relates to a multi-robot collaborative isosurface tracking method under communication restricted conditions. Background Art
[0002] In recent years, the application of robot contour tracking technology in the field of environmental monitoring has developed rapidly, such as forest fire monitoring, marine pollutant monitoring, air pollution monitoring, etc. For large-scale complex tasks, multi-robot collaborative contour tracking has higher efficiency, better robustness and accuracy than single robot methods.
[0003] Isosurface tracking is an extension of field source search: in three-dimensional space, the robot searches for the expected field strength position and tracks the isosurface corresponding to the field strength to obtain the distribution of the signal field and monitor or record the signal field. When the expected field strength is equal to the field source size, isosurface tracking is equivalent to field source search.
[0004] Existing multi-robot isosurface tracking methods often obtain their respective sensor measurements and GPS position information through mutual communication between robots, so as to jointly estimate the gradient information of the signal field and thus achieve tracking of the desired isosurface.
[0005] However, in special extreme environments (such as hostile environments with strong electromagnetic interference, deep-sea environments, etc.), multi-robot systems usually cannot guarantee a stable communication channel, that is, they cannot obtain GPS location information and sensor information of other robots except themselves. Therefore, existing multi-robot isosurface tracking methods cannot be used under communication-restricted conditions.
[0006] Under conditions of limited communication, there are two main difficulties in the multi-robot isosurface tracking method: (1) How can multiple robots achieve collaborative control and efficient collaborative decision-making under conditions of limited communication? (2) The robot motion trajectory is two-dimensional linear, so how to achieve three-dimensional isosurface tracking estimation?
[0007] At present, the common solutions for multi-robot control and decision-making under (1) communication constraints are: a. leader-follower solution; b. intermittent communication solution. The core idea of solution a is to set a leader robot, and all decisions are made locally by the leader robot, while the remaining robots follow the leader robot. This method cannot share effective information between robots without mutual communication, which may lead to low efficiency of decision-making. Solution b requires intermittent communication, allowing robots to actively disconnect communication and re-communicate after a period of time, such as robots in underwater environments regularly floating to the surface for communication. This method adds additional communication costs, limits the flexibility of the system, and the communication information has a strong lag. Summary of the invention
[0008] The purpose of the present invention is to provide a multi-robot collaborative isosurface tracking method under limited communication to address the deficiencies in the prior art, so as to solve the isosurface tracking problem of multiple robots when communication is limited and they cannot exchange sensor measurement values and GPS position information.
[0009] In order to solve the above problems, the present invention provides the following technical solutions:
[0010] The present invention first provides a multi-robot collaborative isosurface tracking method under communication constraints, which comprises the following steps:
[0011] Step 1: Establish a dynamic model of the mobile robot in three-dimensional space:
[0012]
[0013] Among them, r i (t) = [x(t), y(t), z(t)] represents the position of the i-th mobile robot in three-dimensional space at time t; i = 1, 2, ..., N, where N is the number of robots; represents the movement speed of the i-th mobile robot at time t; u i (t) represents the control input of the i-th mobile robot at time t;
[0014] Step 2: Considering the communication limitation problem, all mobile robots detect only through the sensors they carry. The field strength at the robot's own location, the secondary target strength at the robot's own location, and the relative position of the robot itself and all other robots are obtained.
[0015] Step 3: Perform principal component analysis on the relative position information obtained in step 2 to obtain the coordinated motion direction of the multi-robot system without mutual communication.
[0016] Step 4: Design a kinematic controller using the field strength at the robot's own location obtained in step 2, the secondary target strength at the location, and the coordinated motion direction obtained in step 3;
[0017] Step 5: Use the kinematic controller designed in step 4 to collaboratively control multiple robots and achieve the multi-robot isosurface tracking task under limited communication.
[0018] Furthermore, in step 2, the sensor includes a field strength sensor and a distance sensor; the field strength sensor is used to measure the position r of the robot itself at any time. i (t) field strength and secondary target strength; the distance sensor is used to measure the relative position of the robot itself and all other robots.
[0019] Furthermore, the step 2 specifically includes the following sub-steps:
[0020] Step 2.1: Mobile robot i measures its own position r at time t through its own field strength sensor i (t) field strength z i (t), which is expressed as:
[0021]
[0022] in, is the function expression of the signal field to be monitored;
[0023] Step 2.2: Mobile robot i measures its own position r at time t through its own field strength sensor i (t) The secondary target strength f i (t), which is expressed as:
[0024]
[0025] in, is the function expression of the secondary target field to be monitored;
[0026] Step 2.3: Mobile robot i measures its relative position r with all other robots at time t through its own distance sensor ij (t), which is expressed as:
[0027] r ij (t) = r j (t)-r i (t), r ii (t)=0,j=1,2,…,N.
[0028] Furthermore, in the step 2.2, the sub-goal is selected according to the actual application scenario of the robot, and the robot will adaptively optimize the sub-goal on the premise of satisfying the isosurface tracking, and finally converge to the projection of the sub-goal on the isosurface.
[0029] Furthermore, the step 3 specifically includes the following sub-steps:
[0030] Step 3.1: Mobile robot i constructs the corresponding covariance matrix C based on the measured relative position information i (t), which is expressed as:
[0031]
[0032] in The mean of the relative position measured for robot i;
[0033] Step 3.2: Since each robot can observe all other robots, the covariance matrix of all robots is the same C(t), which is expressed in terms of absolute position:
[0034]
[0035] in is the center of all robot positions;
[0036] Step 3.3: Perform eigenvalue decomposition on the covariance matrix obtained in step 3.2 to obtain a set of one-to-one corresponding eigenvalues {λ0,λ1,λ2} and three-dimensional eigenvectors {V0,V1,V2}, where the eigenvectors are sorted according to the size relationship of the eigenvalues λ0<λ1<λ2; the linear combination of the three-dimensional eigenvectors is used as the collaborative motion direction of the multi-robot system.
[0037] Furthermore, the step 4 includes the following sub-steps:
[0038] Step 4.1: The isosurface in the scalar signal field is represented as where z d is the expected field strength;
[0039] Step 4.2: Design the control input of robot i at time t as:
[0040] u i (t) = k0(z i -z d )V0+k1f i V1+k2V2+K c (r c -r i )
[0041] Among them, k0>0, k1>0, k2>0 are control parameters; k c >0, γ>0 is the distance parameter; k0(z i -z d )V0+k2V2 is the isosurface tracking item, which is used to search and track the isosurface, k1f i V1 is the sub-objective optimization item, which is used to optimize the sub-objective at the same time. c (r c -r i ) is the distance control item, which controls the distance between multiple robots by center contraction.
[0042] Compared with the prior art, the present invention has the following beneficial effects:
[0043] (1) The present invention only requires the field strength information and relative position information measured by the robot's own sensors as control input to achieve multi-robot isosurface tracking. It does not require mutual communication and absolute position information between robots and is suitable for environments with restricted communication, such as hostile environments with strong electromagnetic interference, tunnels and underwater environments.
[0044] (2) The present invention proposes a sub-objective optimization method to control the robot to adaptively optimize the sub-objective under the premise of satisfying isosurface tracking. The method has a wide range of applications and can adjust the robot's sub-objective according to actual needs, such as the flight altitude of an aircraft or the diving depth of an underwater submersible.
[0045] (3) The present invention proposes a distance control method for center contraction, which controls the distance only through the relative positions between robots, making it unnecessary to restrict the initial state of the robots and enhancing the robustness of the algorithm. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] Figure 1 It is a schematic diagram of the principle of the method of the present invention.
[0047] Figure 2 The figure is a schematic diagram of the principle of performing principal component analysis on the relative position of the robot in the method of the present invention.
[0048] Figure 3 Schematic diagram of the motion trajectories of multiple robots in an embodiment of the method of the present invention.
[0049] Figure 4 Schematic diagram of field strength sensor measurement values of multiple robots in an embodiment of the method of the present invention.
[0050] Figure 5 Schematic diagram of depth sensor measurement values of multiple robots in an embodiment of the method of the present invention. DETAILED DESCRIPTION
[0051] The implementation of the method of the present invention is described in detail below with reference to the accompanying drawings.
[0052] The present invention proposes a multi-robot collaborative isosurface tracking method under communication constraints. First, a dynamic model of a mobile robot is established in three-dimensional space. Then, the sensor carried by the robot measures the field strength and secondary target value of its own position at the current moment, and measures its relative position with other robots. Then, principal component analysis is performed on the relative position data to obtain the corresponding eigenvector. Finally, a controller is designed based on the above-mentioned measurement values and eigenvectors to control the multi-robots to reach the desired isosurface without communicating with each other, and further track and optimize the secondary target along the isosurface.
[0053] The present invention proposes a multi-robot collaborative isosurface tracking method under communication constraints, the principle of which is as follows Figure 1 As shown, the specific steps include:
[0054] Step 1: Establish a dynamic model of the mobile robot in three-dimensional space:
[0055]
[0056] Among them, r i (t) = [x(t), y(t), z(t)] represents the position of the i-th mobile robot in three-dimensional space at time t; i = 1, 2, ..., N, where N is the number of robots; represents the movement speed of the i-th mobile robot at time t; u i (t) represents the control input executed by the i-th mobile robot at time t.
[0057] Step 2: Considering the communication limitation problem, all mobile robots detect only through their own sensors, which includes the following sub-steps:
[0058] Step 2.1: Mobile robot i measures its own position r at time t through its own field strength sensor i (t), which can be expressed as:
[0059]
[0060] in, is the function expression of the signal field to be monitored.
[0061] Step 2.2: Mobile robot i measures its own position r at time t through its own field strength sensor i The secondary target strength of (t) can be expressed as:
[0062]
[0063] in, is the function expression of the sub-target field to be monitored. The specific sub-target can be selected according to the actual application scenario of the robot, such as the flight altitude of an aircraft or the diving depth of an underwater submersible. The robot will adaptively optimize the sub-target under the premise of satisfying the isosurface tracking, and finally converge to the projection of the sub-target on the isosurface.
[0064] Step 2.3: Mobile robot i measures its relative position to all other robots at time t through its own distance sensor, which can be expressed as:
[0065] r ij (t) = r j (t)-r i (t), r ii (t)=0,j=1,2,…,N
[0066] Step 3: Perform principal component analysis on the relative position information obtained in step 2.3 to obtain the coordinated motion direction of the multi-robot system without mutual communication, which specifically includes the following sub-steps:
[0067] Step 3.1: The mobile robot i constructs the corresponding covariance matrix according to the measured relative position information, which is expressed as:
[0068]
[0069] in The mean relative position measured for robot i.
[0070] Step 3.2: Since each robot can observe all other robots, the covariance matrix of all robots is the same, which can be expressed in terms of absolute position as:
[0071]
[0072] in is the center of all robot positions.
[0073] Step 3.3: Perform eigenvalue decomposition on the covariance matrix obtained in step 3.2 to obtain a set of one-to-one corresponding three-dimensional eigenvectors {V0, V1, V2} and eigenvalues {λ0, λ1, λ2}, such as Figure 2 As shown, the eigenvectors are sorted according to the size relationship of the eigenvalues λ0<λ1<λ2; and the linear combination of the eigenvectors is used as the collaborative motion direction of the multi-robot system.
[0074] Step 4: Design a kinematic controller using the sensor measurements obtained in step 2 and the coordinated motion direction obtained in step 3, including the following sub-steps:
[0075] Step 4.1: The isosurface in the scalar signal field can be expressed as where z d is the expected field strength.
[0076] Step 4.2: Design the kinematic control input of robot i at time t as:
[0077] u i (t) = k0(z i -z d )V0+k1f i V1+k2V2+K c (r c -r i )
[0078] Among them, k0>0, k1>0, k2>0 are control parameters; is the distance controller of center contraction, k c >0, γ>0 is the distance control parameter. k0(z i -z d )V0+k2V2 is the isosurface tracking item, which is used to search and track the isosurface, k1f i V1 is the sub-objective optimization item, which is used to optimize the sub-objective at the same time. c (r c -r i ) is the distance control item, which controls the distance between multiple robots by center contraction.
[0079] Step 5: Use the dynamic controller designed in step 4 to collaboratively control multiple robots and achieve the multi-robot isosurface tracking task under communication constraints.
[0080] The mobile robot and sensor used in the method of the present invention are both conventional equipment; the method of the present invention can be implemented by those skilled in the art through programming.
[0081] The effectiveness of the present invention is verified below in conjunction with a specific implementation.
[0082] Consider a 1km×1km×0.3km three-dimensional underwater environment, where there is an unknown pollution source leaking, and the pollutants diffuse naturally over time. According to existing theories, the natural diffusion process is represented by the following partial differential equation:
[0083]
[0084] in is the pollutant concentration, d = 0.4 km 2 ·h -1 is the diffusion coefficient. Set the desired pollutant concentration z d =3.0ppm, the control parameters of all robots are designed as: k0 = 1.0km·ppm -1 ·h -1 , k1=1.0h -1 , k2=0.2km·h -1 , k c =1.5h -1 , γ = 15m, and the robots do not communicate with each other. In three-dimensional space, two groups of 4 mobile robots are set, and the secondary goals of the first and second groups of robots are set as:
[0085] f 1,i =r i (3)-0.1(km)
[0086]
[0087] That is, the first group of robots is required to track the expected isosurface with an underwater depth of 0.1 km, and the second group of robots is required to track the expected isosurface with an underwater depth of 0.1 km to 0.2 km. The initial positions of the first group of robots are selected as [0.52, 0.30, 0.245], [0.51, 0.30, 0.235], [0.48, 0.30, 0.25], [0.50, 0.30, 0.24] (km), and the initial positions of the second group of robots are selected as [0.13, 0.51, 0.085], [0.13, 0.52, 0.095], [0.13, 0.485, 0.09], [0.13, 0.50, 0.08] (km) to perform the isosurface tracking task. The pollution source starts to spread from time 0, and the robots start to perform the task after 30 minutes. The execution results of the isosurface tracking task are as follows: Figure 3 shown. Figure 3 Schematic diagram of the motion trajectory of a mobile robot in a three-dimensional underwater environment according to an embodiment of the present invention. In the figure, the heat map represents the pollutant concentration, the asterisk represents the initial position of the robot, the diamond represents the current position of the robot, the dotted line represents the first set of robot historical trajectories, and the solid line represents the second set of robot historical trajectories. The tracking error of the robot field strength according to the embodiment of the present invention is shown in FIG. Figure 4 As shown, combined Figure 3 and Figure 4 It can be seen that the pollutants continue to diffuse over time, while the robot always remains on the desired isosurface. Figure 5 As shown in the figure, it can be seen that the robot stably tracks the desired diving depth.
[0088] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims, which all belong to the protection scope of the present invention.
Claims
1. A multi-robot collaborative isosurface tracking method under limited communication, characterized in that: The following steps are involved: Step 1: Establish a dynamic model of the mobile robot in three-dimensional space: Among them, r i (t) = [x(t), y(t), z(t)] represents the position of the i-th mobile robot in three-dimensional space at time t; i = 1, 2, ..., N, where N is the number of robots; represents the movement speed of the i-th mobile robot at time t; u i (t) represents the control input of the i-th mobile robot at time t; Step 2: Considering the communication limitation problem, all mobile robots detect only through the sensors they carry. The field strength at the robot's own location, the secondary target strength at the robot's own location, and the relative position of the robot itself and all other robots are obtained. Step 3: Perform principal component analysis on the relative position information obtained in step 2 to obtain the coordinated motion direction of the multi-robot system without mutual communication. The step 3 specifically includes the following sub-steps: Step 3.1: Mobile robot i constructs the corresponding covariance matrix C based on the measured relative position information i (t), which is expressed as: in The mean of the relative position measured for robot i; Step 3.2: Since each robot can observe all other robots, the covariance matrix of all robots is the same C(t), which is expressed in terms of absolute position: in is the center of all robot positions; Step 3.3: Perform eigenvalue decomposition on the covariance matrix obtained in step 3.2 to obtain a set of one-to-one corresponding eigenvalues {λ0,λ1,λ2} and three-dimensional eigenvectors {V0,V1,V2}, where the eigenvectors are sorted according to the size relationship of the eigenvalues λ0<λ1<λ2; the linear combination of the three-dimensional eigenvectors is used as the collaborative motion direction of the multi-robot system; Step 4: Design a kinematic controller using the field strength at the robot's own location obtained in step 2, the secondary target strength at the location, and the coordinated motion direction obtained in step 3; The step 4 includes the following sub-steps: Step 4.1: The isosurface in the scalar signal field is represented as where z d is the expected field strength; Step 4.2: Design the control input of robot i at time t as: u i (t)=k0(z i -z d )V0+k1f i V1+k2V2+K c (r c -r i ) Among them, k0>0, k1>0, k2>0 are control parameters; k c >0, γ>0 is the distance parameter; k0(z i -z d )V0+k2V2 is the isosurface tracking item, which is used to search and track the isosurface, k1f i V1 is the sub-objective optimization item, which is used to optimize the sub-objective at the same time. c (r c -r i ) is the distance control item, which controls the distance between multiple robots by center contraction; Step 5: Use the kinematic controller designed in step 4 to collaboratively control multiple robots and achieve the multi-robot isosurface tracking task under limited communication.
2. The method for multi-robot collaborative isosurface tracking under limited communication according to claim 1, characterized in that: In step 2, the sensors include a field strength sensor and a distance sensor; the field strength sensor is used to measure the robot's own position r at any time. i (t) field strength and secondary target strength; the distance sensor is used to measure the relative position of the robot itself and all other robots.
3. The method for multi-robot collaborative isosurface tracking under limited communication according to claim 1, characterized in that: The step 2 specifically includes the following sub-steps: Step 2.1: Mobile robot i measures its own position r at time t through its own field strength sensor i (t) field strength z i (t), which is expressed as: in, is the function expression of the signal field to be monitored; Step 2.2: Mobile robot i measures its own position r at time t through its own field strength sensor i (t) The secondary target strength f i (t), which is expressed as: in, is the function expression of the secondary target field to be monitored; Step 2.3: Mobile robot i measures its relative position r with all other robots at time t through its own distance sensor ij (t), which is expressed as: r ij (t)=r j (t)-r i (t),r ii (t)=0,j=1,2,...,N。 4. The method for multi-robot collaborative isosurface tracking under limited communication according to claim 3, characterized in that: In the step 2.2, the sub-goal is selected according to the actual application scenario of the robot. The robot will adaptively optimize the sub-goal on the premise of satisfying the isosurface tracking, and finally converge to the projection of the sub-goal on the isosurface.
