KR-20260064190-A - METHOD FOR PROVIDING SIMULTANEOUS LOCALIZATION AND MAPPING, COMPUTING DEVICE FOR SIMULTANEOUS LOCALIZATION AND MAPPING

KR20260064190AKR 20260064190 AKR20260064190 AKR 20260064190AKR-20260064190-A

Abstract

The present invention relates to a concurrent positioning and mapping method for a computing device, comprising: a step of factoring a joint probability density function of a random finite set from data received from a sensor to the computing device; a step of generating a factor graph based on the factored density function computed by the factoring; a step of updating a set message of the joint probability density function on the factor graph using a set type brief propagation method; and a step of predicting the factor graph based on the update and connecting different variable nodes of the predicted factor graph.

Inventors

김효원

Assignees

충남대학교산학협력단

Dates

Publication Date: 20260507
Application Date: 20241031

Claims (15)

In a simultaneous localization and mapping (SLAM) method for a computing device, A step of factoring the joint probability density function of a random finite set from data received from a sensor to the computing device; A step of generating a factor graph based on a factorization density function calculated by the above factorization; A step of updating the set message of the joint probability density function on the factor graph using a set-type brief propagation method; and The method includes the step of predicting the factor graph based on the above update and connecting different variable nodes of the predicted factor graph. Simultaneous positioning and mapping method for computing devices.
In Article 1, The above-mentioned updating step is, A step of updating a set message propagated from a set factor of the joint probability density function to a set variable; and A step of updating a set message propagated from the set variable of the joint probability density function to the set factor; Simultaneous positioning and mapping method for computing devices.
In Article 2, The above-mentioned updating step is, Step of updating the reliability of the above set variable; and A step of updating the reliability of the above set factor; further comprising Simultaneous positioning and mapping method for computing devices.
In Article 2, The above connecting step is, The step of performing the above prediction and update between arbitrary time k-1 and time k, and Characterized by using a Poisson multi-Bernoulli filter on the set of locus between time k-1 and time k. Simultaneous positioning and mapping method for computing devices.
In Article 4, The above connecting step is, The method further includes the step of shifting an auxiliary variable of a received message within the set message using a conversion factor. Simultaneous positioning and mapping method for computing devices.
In Article 4, The above connecting step is, A further step of inferring the state at time k-1 through smoothing Simultaneous positioning and mapping method for computing devices.
A communication unit that receives data corresponding to the sensing signal of a sensor; A first operation of generating a factor graph based on the above data, a second operation of updating a set message of joint probability density functions on the factor graph, a processing unit that predicts the factor graph based on the update and processes connecting different variable nodes of the predicted factor graph; and A storage unit comprising instructions for processing the first operation, the second operation, and the third operation of the processing unit; Computing device for simultaneous positioning and mapping.
In Article 7, The above processing unit is, A factor graph generation unit that processes the first operation to factorize a joint probability density function of a random finite set from the data by executing the above command, and to generate a factor graph based on the factorization density function calculated by the factorization; A message update unit that processes the second operation to update the set message of the joint probability density function on the factor graph using a set-type reliability propagation method by executing the above command; and A node connection unit comprising: predicting the factor graph based on the update by executing the above command, and processing the third operation to connect different variable nodes of the predicted factor graph. Computing device for simultaneous positioning and mapping.
In Article 8, The above message update unit is, Updating the set message propagated from the set factor of the above joint probability density function to the set variable, and Updating the set message propagated from the set variable of the above joint probability density function to the set factor Computing device for simultaneous positioning and mapping.
In Article 9, The above message update unit is, Update the reliability of the above set variable, and Updating the reliability of the above set factor Computing device for simultaneous positioning and mapping.
In Article 9, The above node connection part is, Characterized by performing the prediction and update between arbitrary times k-1 and k, and using a Poisson multiple Bernoulli filter on the set of trajectories between times k-1 and k. Computing device for simultaneous positioning and mapping.
In Article 11, The above node connection part is, Changing auxiliary variables of the received message within the set message using a conversion factor Computing device for simultaneous positioning and mapping.
In Article 11, The above node connection part is, Inferring the state at time k-1 through smoothing Computing device for simultaneous positioning and mapping.
As a computer-readable recording medium storing a computer program, The above computer program is, Includes instructions for enabling a processor to perform concurrent positioning and mapping methods of a computing device, and The above method is, A step of factoring the joint probability density function of a random finite set from data received from the sensor to the computing device; A step of generating a factor graph based on the factorization density function calculated by the above factorization; A step of updating a set message of the joint probability density function on the factor graph using a set-type reliability propagation method; and The method includes the step of predicting the factor graph based on the above update and connecting different variable nodes of the predicted factor graph. Computer-readable recording medium.
As a computer program stored on a computer-readable recording medium, The above computer program is, Includes instructions for enabling a processor to perform concurrent positioning and mapping methods of a computing device, and The above method is, A step of factoring the joint probability density function of a random finite set from data received from the sensor to the computing device; A step of generating a factor graph based on the factorization density function calculated by the above factorization; A step of updating a set message of the joint probability density function on the factor graph using a set-type reliability propagation method; and The method includes the step of predicting the factor graph based on the above update and connecting different variable nodes of the predicted factor graph. A computer program stored on a recording medium.

Description

Method for providing simultaneous localization and mapping of a computing device, computing device for simultaneous localization and mapping The present invention relates to a simultaneous localization and mapping (SLAM) technique for computing devices. Joint probability density refers to the probability density of two or more continuous probability variables, that is, the probability density of two occurring simultaneously, and uses vectors or scalars as variables. If such joint probability densities can be factorized, they can be expressed as factored probability density functions and represented as factor graphs, and the marginal probability density of each variable can be efficiently calculated on these factor graphs. However, when the variable is a set with an undetermined number of elements, there is a problem in that the probability density of the factored set cannot be represented as a factor graph. The aforementioned background technology is technical information that the inventor possessed for the derivation of the present invention or acquired during the process of deriving the present invention, and it cannot be considered as publicly known technology disclosed to the general public prior to the filing of the present invention. FIG. 1 is a schematic block diagram illustrating the functions of a computing device for simultaneous positioning and mapping according to an embodiment of the present invention. Figure 2 is a block diagram illustrating the detailed functions of the processing unit of the computing device of Figure 1. Figure 3 is a diagram illustrating, in an exemplary manner, the factor graph of the set density provided by the factor graph generation unit of Figure 2. Figure 4 is a diagram illustrating an exemplary Poisson multi-Bernoulli (PMB) density function having auxiliary variables provided by the factor graph generation unit of Figure 2. FIG. 5 is a diagram that exemplarily illustrates the process of connecting variable nodes based on updates processed in the node connection part of FIG. 2. The advantages and features of the present invention and the methods for achieving them will become clear by referring to the embodiments described below in detail together with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below but can be implemented in various forms. These embodiments are provided merely to ensure that the disclosure of the present invention is complete and to fully inform those skilled in the art of the scope of the invention, and the scope of the present invention is defined only by the claims. In describing the embodiments of the present invention, specific descriptions of known functions or configurations will be omitted unless actually necessary for describing the embodiments of the present invention. Furthermore, the terms described below are defined in consideration of the functions in the embodiments of the present invention, and these may vary depending on the intentions or practices of the user or operator. Therefore, such definitions should be based on the content throughout this specification. Reliability propagation is a probabilistic inference algorithm useful for efficiently calculating the approximate limit probability density of random variables. However, while reliability propagation in its standard form applies only to vector-type random variables with a fixed and known number of vector factors, certain applications tend to rely on random finite sets (RFS) with an unknown number of vector factors. In order to derive a new inference method for each random finite set, the present invention proposes a reliability propagation rule for a factor graph defined on a sequence of random finite sets, each containing an unknown number of factors. Furthermore, vector-based reliability propagation can be specifically applied to collective reliability propagation, where each random finite set follows a Bernoulli process. To validate the proposed collective reliability propagation, a PMB filter for simultaneous localization and mapping (SLAM) is applied, which naturally leads to a collective reliability propagation PMB-SLAM technique similar to the vector-based SLAM method. In particular, reliability propagation techniques, which are utilized as statistical inference algorithms in various fields, are effectively used to calculate probability densities in Bayesian networks, where, given observed data, the marginal probability density of a random variable of interest can be efficiently calculated. Reliability propagation is being more actively applied to mapping, multiple-target tracking (MTT), SLAM, and simultaneous localization and tracking (SLAT). In an embodiment of the present invention, we propose a concurrent positioning and mapping technology for a computing device suitable for computing a joint probability density function based on reliability propagation for input data, factoring the joint probability density