A Network Cooperative Localization Method and Device Based on Momentum Acceleration

Abstract

The present invention discloses a network cooperative positioning method and device based on momentum acceleration. The method includes: establishing a cooperative positioning network including anchor nodes and nodes to be positioned, initializing the positions of the nodes to be positioned in the network, and establishing communication links between adjacent nodes to be positioned through ultra-wideband communication methods; each node to be positioned obtains noisy ranging information with adjacent anchor nodes and other nodes to be positioned; using the obtained noisy ranging information, updating the position coordinates of each node to be positioned by using the gradient descent method based on momentum acceleration until a preset number of iterations or the result converges, and completing the positioning; each node to be positioned in the positioning device includes an observation acquisition module, a positioning calculation module, an information interaction module, and a result display module. The present invention can overcome the defects of the existing technology and solve the problem of poor real-time performance of UWB network cooperative positioning.

Description

Technical Field

[0001] The present invention belongs to the technical field of positioning, and particularly relates to a network cooperative positioning method and device based on momentum acceleration. Background Art

[0002] Current indoor positioning technologies include WIFI (wireless hotspot) positioning, Bluetooth positioning, RFID (radio frequency identification) positioning, UWB (ultra-wideband) positioning, etc. Among them, UWB positioning has higher positioning accuracy compared with other positioning technologies. Due to its ultra-wideband characteristics, its positioning accuracy can reach the centimeter level.

[0003] In practical applications, obtaining high-precision position information usually requires a large number of deployed reference nodes. Compared with positioning technologies such as WIFI, Bluetooth, and RFID, the deployment cost of reference nodes in UWB positioning is much higher. Therefore, reducing the deployment of UWB reference nodes can effectively reduce the cost of the UWB positioning system. If high-precision position information is to be obtained in an environment with a small number of deployed UWB reference nodes, it can be achieved through multi-node network cooperative positioning technology. In existing cooperative positioning methods, it is usually necessary to perform multiple information exchanges and iterative processes between nodes to obtain the position information of each node. In a large-scale cooperative network, a large number of iterative times and information exchanges are usually required. A large number of iterations and information exchanges make the communication overhead of the network very large, and at the same time affect the real-time performance of the entire network positioning. Therefore, an accelerated cooperative positioning method becomes very important in real-time positioning and applications. Summary of the Invention

[0004] In view of this, the present invention provides a network cooperative positioning method and device based on momentum acceleration, which can overcome the defects of the existing technology and solve the problem of poor real-time performance of UWB network cooperative positioning.

[0005] The technical solution of the present invention is implemented as follows:

[0006] A network cooperative positioning method based on momentum acceleration, comprising the following steps:

[0007] Step 1, establish a cooperative positioning network including anchor nodes and nodes to be located, initialize the positions of each node to be located in the network using prior information or randomly within the region, and establish communication links between adjacent nodes to be located through ultra-wideband communication methods;

[0008] Step 2, each node to be located obtains the position information of adjacent anchor nodes, and obtains noisy ranging information between each node to be located and adjacent anchor nodes as well as other nodes to be located through UWB technology;

[0009] Step 3: Each node to be located exchanges the currently estimated position information with adjacent nodes to be located, and uses the obtained noisy ranging information to update the estimated position coordinates of each node to be located using the gradient descent method with momentum acceleration;

[0010] Step 4: Repeat Step 3 until a preset number of iterations or the result converges to complete the positioning.

[0011] Further, the gradient descent method with momentum acceleration is specifically as follows:

[0012] During the first iteration of the position of the node to be located, the position gradient at the node to be located is obtained using the initial position information of the cooperative positioning network and the noisy ranging information between nodes, and then the position coordinates of the node to be located are updated according to the position gradient; starting from the second iteration until the preset number of iterations or the result converges, based on the current position estimation coordinates obtained after the nodes exchange position information, the position gradient is calculated, and then the momentum acceleration gradient is calculated in combination with the position gradient of the previous iteration, and finally the position coordinates are updated according to the momentum acceleration gradient.

[0013] Further, the gradient descent method with momentum acceleration is specifically as follows:

[0014] In the network cooperative positioning using ranging, the gradient estimation i at the position θ of Agent i can be expressed as:

[0015]

[0016]

[0017] where A i represents the set of Agents adjacent to Agent i in the cooperative network, B i represents the set of Anchors adjacent to Agent i; r ij and r im respectively represent the noisy ranging to the adjacent Agent j and Anchor m; θ i represents the current position coordinate vector of Agent i, θ j and θ m respectively represent the position coordinate vectors of the adjacent Agent j and Anchor m to Agent i; e ij and e im respectively represent the unit direction vectors from Agent i to the adjacent Agent j and Anchor m;

[0018] Then, based on the gradient of Agent i, the momentum acceleration gradient The specific formula is as follows:

[0019]

[0020] Among them, l is the number of iterations, and β is the attenuation coefficient;

[0021] Finally, use the iterative formula to calculate the position estimation coordinates The specific formula is:

[0022]

[0023] Among them, α is the iterative step size.

[0024] Furthermore, the position coordinates of each node to be located are updated in a synchronous or asynchronous manner.

[0025] A network collaborative positioning device based on momentum acceleration, each node to be located includes an observation acquisition module, a positioning calculation module, an information interaction module and a result display module, where:

[0026] The observation acquisition module is used to obtain the initial position coordinates of the current node to be located and obtain the noisy distance information between the current node to be located and its adjacent nodes. The initial position coordinates can be set as prior position information or random coordinates within the network area;

[0027] The positioning calculation module is used to update the position coordinates of the current node to be located using the obtained noisy distance information, the initial position coordinates of the node to be located and its adjacent nodes to be located, and the position coordinates of the anchor nodes, using the gradient descent method based on momentum acceleration;

[0028] The information interaction module is used to realize the position information interaction between the current node to be located and its adjacent nodes to be located and send the position information to the result display module;

[0029] The result display module is used to judge whether the number of update iterations of the position of the current node to be located reaches the preset number or the result converges. If so, the current position coordinates are determined as the final position coordinates of the node to be located, and a mark is displayed on the network area map.

[0030] A network collaborative positioning device system based on momentum acceleration, including UWB nodes, a processor, a memory, Bluetooth nodes and a mobile display terminal; among them, the UWB nodes are responsible for obtaining noisy ranging information and exchanging position information with other UWB nodes to be located; the processor is responsible for executing the positioning algorithm program stored in the memory, that is, the network collaborative positioning method based on momentum acceleration; the memory is used to store the positioning algorithm program; the Bluetooth nodes are used to send the calculation results of the processor to the mobile display terminal through Bluetooth; the mobile display terminal is used to display the plane map of the current positioning environment and display the UWB node positioning results sent by the marked Bluetooth nodes.

[0031] Beneficial effects:

[0032] 1. Compared with the existing technology, since the momentum acceleration gradient is used in the positioning method provided by the present invention, it can achieve cooperative positioning under UWB conditions faster than the traditional gradient descent method, and has the effect of improving the real-time performance of the UWB cooperative positioning system while ensuring the positioning accuracy.

[0033] 2. Compared with the existing devices, the positioning device provided by the present invention has a simple structure, and the required equipment is easy to obtain, making the assembly of the device easier to achieve.

[0034] 3. The display function provided by the device of the present invention can provide visual positioning results to the users of the device, which is convenient for users to use.

[0035] 4. The present invention also provides a low-cost positioning device implementation solution, and designs a low-cost distributed time-division broadcast polling networking scheme based on UWB. Description of the drawings

[0036] Figure 1 It is a schematic flow chart of a network cooperative positioning method based on momentum acceleration provided by the present invention.

[0037] Figure 2 It is a comparison diagram of the simulation effects of a network cooperative positioning method based on momentum acceleration provided by the present invention.

[0038] Figure 3 It is a schematic structural diagram of a network cooperative positioning device based on momentum acceleration provided by the present invention.

[0039] Figure 4 It is a schematic diagram of a construction scheme of a set of visual portable network cooperative positioning device system provided by the present invention.

[0040] Figure 5 It is a schematic diagram of the specific structure of each user node in a construction scheme of a set of visual portable network cooperative positioning device system provided by the present invention. Detailed implementation manners

[0041] The following takes examples in conjunction with the drawings and describes the present invention in detail.

[0042] The present invention provides a cooperative positioning method based on momentum acceleration, as Figure 1 shown, and its specific steps include:

[0043] Step 1: Establish a cooperative positioning network containing M Anchor nodes and N Agent nodes to be located, initialize the positions of the nodes in the positioning network, and establish communication links between nodes through ultra-wideband communication methods. In the positioning network, the positions of the Anchor nodes are preset positions set artificially, and the initial position coordinates of the nodes to be located can be set as prior position information or random coordinates within the network area.

[0044] Step 2: Each node to be located obtains noisy ranging information with adjacent Anchor nodes and other nodes to be located. Among them, adjacent nodes are defined as nodes with a communication radius less than a set threshold or signal strength less than a set threshold with the current node to be located.

[0045] Step 3: Use the obtained noisy ranging information and use the gradient descent method based on momentum acceleration to update the position coordinates of each node to be located until the preset number of iterations or the result converges, and complete the positioning. The gradient descent update method based on momentum acceleration is as follows:

[0046] In network cooperative positioning using ranging, the gradient estimate of Agent i at position θ i can be expressed as:

[0047]

[0048]

[0049] where A i represents the set of Agents adjacent to Agent i in the cooperative network, and B i represents the set of Anchors adjacent to Agent i; r ij and r im respectively represent the noisy ranging with adjacent Agent j and Anchor m; θ i represents the current position coordinate vector of Agent i, θ j and θ m respectively represent the position coordinate vectors of adjacent Agent j and Anchor m with Agent i; e ij and e im respectively represent the unit direction vectors from Agent i to adjacent Agent j and Anchor m.

[0050] Then, according to the gradient of Agent i, the momentum acceleration gradient at the l-th iteration is estimated The specific formula is as follows:

[0051] ​

[0052] Among them, l is the number of iterations, and β is the attenuation coefficient.

[0053] Finally, the position estimation coordinates are obtained using the iterative formula The specific formula is:

[0054]

[0055] Among them, α is the iterative step size.

[0056] The specific algorithm comparison results are shown by Figure 2 As shown. Compared with the traditional gradient descent method, the method of the present invention requires fewer iterations under the condition of achieving the same positioning accuracy. When using UWB for cooperative positioning, better real-time performance can be guaranteed. In particular, in conventional accelerated learning algorithms, the values of parameters such as α and β are usually constants. The method of the present invention uses a method of automatically adjusting according to the number of iterations to obtain the values of parameters α and β, and obtains an iterative speed faster than the conventional value-taking method, and can obtain the positioning result with fewer times under the condition of ensuring the positioning accuracy.

[0057] Corresponding to the above method example, the present invention example provides a cooperative positioning device based on momentum acceleration, as Figure 3 shown. Each node to be positioned in the device includes an observation acquisition module, a positioning calculation module, an information interaction module, and a result display module, where:

[0058] The observation acquisition module is used to obtain the initial position coordinates of the current node to be positioned and obtain the noisy ranging information between the current node to be positioned and its adjacent nodes. The initial position coordinates can be set as prior position information or random coordinates within the network area; the positioning calculation module is used to update the position coordinates of the current node to be positioned using the obtained noisy ranging information, the initial position coordinates of the node to be positioned and its adjacent nodes to be positioned, and the position coordinates of the anchor nodes, using the gradient descent method based on momentum acceleration. The information interaction module is used to realize the position information interaction between the current node to be positioned and its adjacent nodes to be positioned and send the position information to the result display module; the result display module is used to judge whether the number of update iterations of the position of the current node to be positioned reaches the preset number or the result converges. If so, the current position coordinates are determined as the final position coordinates of the node to be positioned and are marked and displayed on the network area map.

[0059] The embodiment of the present invention also provides a set of positioning device system construction solutions. Among them, the scenario applied by the solution is as Figure 4 shown. In a cooperative positioning network, it includes several Anchors and user nodes. Adjacent nodes in the network can communicate and range with each other. The specific structure of each user node is as Figure 5As shown in the figure, it includes a UWB node, a processor, a memory, a Bluetooth node, and a mobile display terminal. Among them, the UWB node is responsible for obtaining ranging information and position update information of other UWB nodes to be located, and sending its own position update information to other UWB nodes to be located; the processor is responsible for executing the program stored in the memory to implement a cooperative positioning method based on momentum acceleration provided by the embodiments of the present invention; the memory is used to store computer programs; the Bluetooth node is used to send the calculation results of the processor to the mobile display terminal through Bluetooth; the mobile display terminal is used to display the plane map of the current positioning environment and the positioning results of the UWB nodes sent by the Bluetooth node.

[0060] A set of positioning device system construction solutions provided by the embodiments of the present invention can be held and used by users, and the positioning results are displayed in real time on the mobile display terminal, with good portability and visualization effects. At the same time, taking advantage of the feature that the UWB node itself can send messages, this solution designs and implements a low-cost cooperative networking solution based on UWB, that is, a distributed time-division broadcast polling networking solution. This solution pre-numbers the Agents in the cooperative network as [N1, N2,... N n . When the cooperative network starts running, it first starts from the Agent numbered N1, uses the UWB signal to sequentially obtain the noisy ranging information between other nodes in the network, and determines which adjacent nodes of the current Agent are based on the threshold. Then, after updating its own position coordinates using the momentum acceleration algorithm provided by the embodiments of the present invention, the Agents in the cooperative network broadcast the position update information to other UWB nodes to be located in the network in sequence according to the number, and add their own tag information to the sent information. Finally, after receiving the message broadcast, the Agents in the network determine whether the message source is sent from an adjacent node according to the tag information attached to the message, and continue to update their own position coordinates using the message sent by the adjacent node. The entire cooperative network repeats the above process until the preset number of iterations is reached.

[0061] In summary, the above is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A network collaborative positioning method based on momentum acceleration, characterized in that It includes the following steps: Step 1: Establish a cooperative positioning network including anchor nodes and nodes to be located. Initialize the positions of each node to be located in the network using prior information or randomly within the area, and establish communication links between adjacent nodes to be located through ultra-wideband communication methods; Step 2: Each node to be located obtains the position information of adjacent anchor nodes, and obtains noisy ranging information between each node to be located and adjacent anchor nodes as well as other nodes to be located through UWB technology; Step 3: Each node to be located exchanges the currently estimated position information with adjacent nodes to be located, and uses the obtained noisy ranging information to update the estimated position coordinates of each node to be located using the gradient descent method with momentum acceleration; The gradient descent method with momentum acceleration is specifically as follows: In network cooperative positioning using ranging, the Agent i at the position θ i gradient estimation can be expressed as: Among them, A i represents the set of Agents adjacent to Agent i in the cooperation network, and B i represents the set of Anchors adjacent to Agent i ; r ij and r im respectively represent the noisy ranging to the adjacent Agent j and Anchor m ; represents the current position coordinate vector of Agent i ; θ j and θ m respectively represent the position coordinate vectors of the Agent i adjacent to Agent j and Anchor m ; e ij and e im respectively represent the unit direction vectors from Agent i to the adjacent Agent j and Anchor m ; Next, according to the Agent i , the l momentum-accelerated gradient of the th iteration is estimated, and the specific formula is as follows: Among them, l is the number of iterations, β is the attenuation coefficient; Finally, use the iterative formula to calculate the position estimation coordinates , and the specific formula is as follows: Among them, α is the iteration step size; Step 4: Repeat Step 3 until the preset number of iterations or the result converges to complete the positioning.

2. The network collaborative positioning method based on momentum acceleration according to claim 1, wherein The position coordinates of each node to be located are updated in a synchronous or asynchronous manner.

3. A network collaborative positioning device based on momentum acceleration that applies the method described in claim 1, characterized in that, Each node to be located includes an observation acquisition module, a positioning calculation module, an information interaction module, and a result display module, where: The observation acquisition module is used to obtain the initial position coordinates of the current node to be located and obtain the noisy distance information between the current node to be located and its adjacent nodes. The initial position coordinates can be set as prior position information or random coordinates within the network area; The positioning calculation module is used to update the position coordinates of the current node to be located using the obtained noisy distance information, the initial position coordinates of the node to be located and its adjacent nodes to be located, and the position coordinates of the anchor nodes, using the gradient descent method with momentum acceleration; The information interaction module is used to realize the position information interaction between the current node to be located and adjacent nodes to be located and send the position information to the result display module; The result display module is used to judge whether the number of update iterations of the position of the current node to be located reaches the preset number or the result converges. If so, the current position coordinates are determined as the final position coordinates of the node to be located, and a label is displayed on the network area map.

4. A network collaborative positioning device system based on momentum acceleration that applies the method described in claim 1, characterized in that, It includes UWB nodes, a processor, a memory, Bluetooth nodes, and a mobile display terminal; among them, the UWB nodes are responsible for obtaining noisy ranging information and exchanging position information with other UWB nodes to be located; the processor is responsible for executing the positioning algorithm program stored in the memory, and the memory is used to store the positioning algorithm program; the Bluetooth nodes are used to send the calculation results of the processor to the mobile display terminal through Bluetooth; the mobile display terminal is used to display the plane map of the current positioning environment and display the positioning results of the UWB nodes sent by the labeled Bluetooth nodes.

Images