Robot positioning, sensor network calibration and environmental mapping are the three basic problems of mutual coupling in the universal robot system. The effective solution is the premise that the universal robot system provides efficient and intelligent services. This paper proposes the concept of simultaneous robot positioning, sensor network calibration and environmental mapping for ubiquitous robot systems. By analyzing the coupling relationship between the three, the joint conditional probability representation of simultaneous positioning, calibration and mapping problems is given. The formula and Markov characteristics decompose it into several solvable terms, and use the idea of â€‹â€‹Rao-Blackwellized particle filter to solve them separately. Firstly, the joint sensor network observes the robot, the robot observes the characteristics of the positioned environment and the robot's own control quantity, and designs the sampling proposal distribution and weight update formula of the pose particle. Secondly, the joint sensor network has the robot motion track and The recursive formula of sensor network calibration is designed by observing the characteristics of the environment. Then, the recursive formula of environmental mapping is designed by joint sensor network and robot observation of environmental characteristics (positioned or newly discovered). The complete simultaneous positioning, calibration and mapping algorithms are given, and the effectiveness of the algorithm is verified by simulation experiments.
In the field of robotics research, the universal robot system is the intersection of sensor network technology and service robot technology [1âˆ’3]. By combining these two technologies effectively, the advantages of the two are complemented: on the one hand, the sensor network is visible. For the extension of the robot's perception capability, the sensor network distributed throughout the environment can provide the robot with global sensing capability to compensate for the defects of the robot's global environment perception capability. On the other hand, the robot can be regarded as the actuator of the sensor network, thus enabling the sensor network to have Active service capabilities. Therefore, the universal robot system has important application prospects in service environments with large area and strong dynamics, such as homes, hospitals, and exhibition halls.
As we all know, self-positioning and environmental map construction is the basis for mobile robots to carry out environmental cognition and path planning, and ultimately provide efficient and intelligent services . For ubiquitous robot systems, sensor network calibration is also an important part of ensuring its efficient operation. The calibrated sensor network can locate robots and environmental dynamic targets in real time, and dynamically update the environment map accordingly. In fact, the data collected without knowing the location of the sensor nodes does not make much sense in practical applications . At the beginning of introducing the universal robot system into a certain service environment, robot positioning, sensor network calibration and environmental mapping have become the three basic problems it faces . The positioning is continuous throughout the robot's work, and the sensor network calibration and environmental map construction are completed at the beginning of the normal operation of the universal robot system, and are updated in real time as the network node or environment changes during the work. In addition, the analysis shows that the three problems of robot positioning, sensor network calibration and environmental mapping in the universal robot system are relatively independent and coupled with each other: on the one hand, the sensor network provides global observation, which can assist the robot to complete the positioning under dynamic environment; On the one hand, accurate calibration of the sensor network is a prerequisite for its assisted robot positioning. Furthermore, the observation of the built map will help to improve the calibration accuracy of the sensor network and the accuracy of the robot positioning. If we only discuss the coupling relationship between robot positioning and environmental mapping, it will degenerate into the problem of Simultaneous localizaTIonandmapping (SLAM), which is widely studied by scholars at home and abroad.
In terms of simultaneous robot positioning and environmental mapping, the use of probabilistic methods to solve the SLAM problem is the current research direction and hotspots . Among them, the extended Kalman filter (Extended Kalman? l-ter, EKF) SLAM method is not suitable for solving the existence. The estimation problem of non-Gaussian noise conditions ; the particle filtering method is suitable for nonlinear and non-Gaussian cases, but this method is too computationally expensive when the problem dimension is high, and it is difficult to meet the real-time requirements of the system; Rao-Blackwellized particles The filtering uses the extended Kalman filter to process the nonlinear part and the particle filter to deal with the non-Gaussian part. Therefore, it has the advantages of extended Kalman filtering and particle filtering, and has been successfully applied in SLAM [6âˆ’8]. However, due to the airborne sensor, the robot observation error is coupled with the motion error, and the positioning and mapping errors will inevitably spread with the robot's moving distance . In , a particle-assisted wireless sensor network-assisted SLAM method based on particle filter is proposed to solve the problem of high spatial dimension and multi-data association problem. The literature  proposes a mobile robot SLAM method based on distributed sensing of sensor networks. To create accurate maps of large-scale environments;  proposed a particle-filter-based mobile robot SLAM algorithm in wireless sensor network environment. Although the above methods all reflect the idea of â€‹â€‹using the global observation of the sensor network to assist the SLAM, but the calibration problem of the sensor network node has not been considered, and the effective solution of the problem is the basis of the sensor network to assist the robot positioning and environmental feature mapping.
In terms of simultaneous sensor network calibration and robot positioning, the literature  proposed a robot-based camera network online self-calibration method. Based on this, the literature  proposed a distributed perceptual cooperative mobile robot Monte Carlo positioning. The method utilizes observations of the calibration sensor network to assist in robot positioning. In fact, similar to the SLAM problem, sensor network calibration and robot positioning can be performed simultaneously (ie, simultaneous sensor network calibration and robot positioning). For this problem, the literature  proposed a simultaneous sensor network based on Rao-Blackwellized particle filter. Calibration and robot positioning methods, but the research on joint robot positioning, sensor network calibration and environmental mapping has not yet been launched.
Considering the coupling relationship between the robot positioning, the sensor network calibration and the environmental mapping problem in the ubiquitous robot system, in order to fully integrate the multiple types of information sources involved in the positioning, calibration and mapping process, and avoid the cumbersome offline calibration of the sensor network, This paper proposes the concept of simultaneous positioning, calibration and mapping of universal robot systems. In the joint solution of the three, theoretical analysis is carried out from the perspective of probability, and the joint conditional probability is decomposed into several solvable terms to solve separately. Based on the Rao-Blackwellized particle filter idea, combined robot control information, sensor network observation of the robot, and robot observation of the built environment map to estimate the robot pose particle and its weight distribution, and then according to the sensor network for the robot and the built environment The observation of the map to calibrate the parameters of the sensor network, and finally the robot and sensor network to observe the environment to construct a feature map of the home environment.
1 system description
1.1 system composition
The object discussed in this paper is the ubiquitous robotic system, which consists of three parts: a sensor network with universal sensing and processing capabilities, a processing host, a mobile service robot interacting with a ubiquitous processing host, and each in the environment. Targets (including service objects, operational items, and environmental roadmaps, etc., collectively referred to herein as targets). Figure 1 shows a schematic diagram of a typical home ubiquitous robotic system. The implementation scheme is as follows: The RGB-D camera is used as a node to construct a sensor network. The camera can simultaneously acquire the color and distance information of the target in the field of view. The -D camera is connected to an image capture card of a processing host through a data line. The host is responsible for analyzing and processing the images captured by the cameras, and communicating with the service robot through the wireless network; the home service robot platform is equipped with a hand-eye system. Pioneer3DX mobile robot, and designing color blocks for the robot to observe and locate the sensor nodes; selecting the scale-variant feature transform (SIFT) of the home environment and target  for feature detection, matching and recognition. In addition, the QRcode tag that identifies the name, function, and usage of the common target for the family can achieve deeper knowledge of the item by reading the tag robot.
Node deployment of sensor networks needs to consider factors such as node observation range, energy consumption, and obstacle distribution. This problem has been discussed in related literature [17âˆ’19]. It is not discussed here, assuming sensor nodes have been deployed reasonably. .
2Universal robot system simultaneous positioning, calibration and mapping
2.1 basic ideas
From the point of view of probability, the probability density of the universal robot system for simultaneous positioning, calibration and mapping can be used.
We provide wire harness manufacturing services for cable harnesses and built-to-print cables used in many industries, such as computer, game machine, POS machine, ATM , audio/video, electro-mechanical, data communications, telecommunications, medical, etc.
Related Products:speaker cable,customized cable assemblies,electrical cable assemblies.
We have developed a tradition of high-tech engineering, prototyping, and quality custom cable manufacturing at very competitive pricing. Also with professional flow chart (wire cutting-stripping-copper twisting-crimping-crimping 100% inspection-soldering-molding-asssembling-testing-FQC100% -OQC) , which can help us support customers with stable quality.
Speaker Cable,Straight Bnc Cable Assembly,Idc Red Custom Cable Assembly,Customized Cable Assemblies,EV Cable Assemblies,Cable Assembly,Custom Cable,Customized Cable Assemblies
ETOP WIREHARNESS LIMITED , http://www.oemmoldedcables.com