CN107091642A - A kind of indoor orientation method based on the mapping of different plane anchor node and rasterizing correction - Google Patents

Abstract

The invention discloses an indoor positioning method based on different-plane anchor node mapping and grid deviation correction, which includes the following three steps: (a) establishing a different-plane test model; (b) dimensionality reduction processing of the projection plane; (c) Same plane precision adjustment for skew correction. Mainly aiming at the problem of inaccurate positioning and path planning in the case of indoor multiple obstacles, an indoor positioning method is proposed, which maps the obstacle projection onto the same plane for dimensionality reduction processing, and performs a deviation correction mode for planar map rasterization. The present invention The method simplifies the existing non-line-of-sight measurement and positioning of a large amount of communication overhead, improves the positioning accuracy of a mobile object that needs to be positioned in a multi-obstacle environment, and improves the reliability of path planning.

Description

An Indoor Positioning Method Based on Out-of-Plane Anchor Node Mapping and Rasterization Correction

technical field

The invention belongs to a positioning technology for reducing positioning errors in an indoor environment, in particular to an indoor positioning method based on different-plane anchor node mapping and grid deviation correction.

Background technique

With the improvement of living standards and the advancement of science and technology, many smart home products continue to enter our lives, bringing us great convenience and greatly improving our quality of life. Service robots are products that are born in response to the society and people's pursuit of high-quality and easy life. In China and even countries around the world, home service robots have a huge customer base and market, and have become an important development trend in the development of the robot industry. A key direction of robot development.

During the development of service-oriented intelligent robots, general robots or other smart home service mobility products are mainly used indoors, so positioning and path planning technologies directly affect the intelligence and efficiency of services. For indoor positioning technologies, the existing There are seven positioning technologies, namely infrared positioning technology, ultrasonic indoor positioning technology, radio frequency identification (RFID) indoor positioning technology, Bluetooth indoor positioning technology, Wi-Fi indoor positioning technology, ZigBee indoor positioning technology and ultra-wideband indoor positioning technology. In terms of indoor mobile positioning of robots and other smart products, the main solution is location information and path planning to the destination. In the indoor situation, due to the occlusion of buildings and multipath propagation effects, the positioning effect is not ideal. Therefore, indoor positioning is also a hot issue in research and positioning. There are mainly two types of positioning used in existing indoor positioning: ranged-based indoor positioning and range-free indoor positioning. The former needs to know the distance or angle information between beacon nodes, while the latter mainly relies on the adjacency and connectivity between nodes. The ranging-based indoor positioning method and algorithm calculate the position of the location node according to the measured distance or angle information between nodes, mainly based on TOA, based on time difference of arrival (Time Difference of Arrival, TDOA), based on angle of arrival ( Angle of Arrival, AOA) and positioning algorithms based on RSSI values. The commonly used ranging algorithm models are as follows: trilateration, triangulation, and maximum likelihood estimation. The positioning algorithm based on non-ranging does not need to use information such as coordinates, distances, and angles between nodes, and generally uses the network connectivity to perform preliminary positioning, which realizes low-cost indoor positioning, but the positioning error is usually lower than that of ranging algorithms. big. Commonly used non-ranging positioning algorithms include: centroid algorithm, DV-Hop (Distance Vector Routing, DV), convex programming algorithm, fingerprint-based RSSI algorithm, approximate triangle interior point measurement algorithm (Approximately Point-In-Triangulation, APIT) .

There is a certain difference between the working environment of the existing positioning research and the actual working environment of the home service robot. The particularity of the working environment of the home service robot is not considered. Due to the influence of a large number of walls, doors and soft decorations in the home, the positioning There is often a certain deviation between the location information and the actual situation, and the signal is also affected by factors such as multiple diffraction, absorption, and multipath during the propagation process. The existence of these factors causes the home robot to be unable to accurately obtain its own location information and the location information of the target point, thereby affecting whether the home service robot can continuously serve across multiple sub-working areas. This requires a more suitable positioning method for home service robots based on the particularity of the home environment.

Contents of the invention

Purpose of the invention: Aiming at the problem of the above-mentioned mobile positioning accuracy of the existing home service robot under the interference of a large number of walls and doors, and the indoor positioning using distance measurement based on the high communication overhead and high cost required for high-precision positioning, the present invention proposes a An indoor positioning method based on different plane anchor node mapping and grid deviation correction, which is suitable for recalculating the positioning information of the robot in the home environment and improving the accuracy of the trilateration positioning algorithm in the indoor NLOS environment.

Technical solution: An indoor positioning method based on different-plane anchor node mapping and grid deviation correction described in the present invention includes the following three steps:

(a) Establish a different plane test model;

(b) Dimensionality reduction processing of the projected plane;

(c) Correct the deviation with the accuracy of the same plane and obtain the coordinates of the point to be positioned.

Wherein, the establishment of the different-plane test model in the step (a) includes the rasterization of the environment map and the location arrangement of indoor anchor nodes: the rasterization of the environment map is to vertically project the indoor obstacles onto the same plane to obtain the indoor Plane map, and then perform expansion processing on the size of the obstacle mapped to the indoor planar map, and then uniformly rasterize the indoor planar map to obtain a grid map, and divide the grid map into obstacles according to the mapping of obstacles area, free area and dynamic area, and then the grid map is numbered sequentially; the indoor anchor node layout is to set the iBeacon beacon node on the top of the roof, and then measure the position of the anchor point respectively The linear distance from each anchor node and the vertical height of each iBeacon beacon node, and the RSSI value at different distances between the positioning point and iBeacon were measured and collected more than twice, and the RSSI and the distance curve between the positioning point and the anchor node were fitted by Matlab. Establish an indoor attenuation model that simulates a suitable home environment, and obtain the RSSI and the distance relationship between the anchor point and the anchor node.

The obstacle area is a fixed object, including indoor walls and static objects; the free area is an obstacle-free area; the dynamic area includes a door and a projected location area of movable objects.

In the step (b) dimension reduction of the projected plane, the indoor attenuation model is used to convert the obtained RSSI value of the anchor point and the anchor node into distance information, and at the same time, the distance is projected to a two-dimensional plane through the Pythagorean theorem to obtain The horizontal straight-line distance between the positioning point and the anchor node; then through the ranging algorithm and distance information D_actual, the estimated position of the point to be positioned is obtained.

The step (c) is to adjust the deviation and rectify the accuracy of the same plane according to the estimated position obtained in step (b) to calculate the position information and D_virtual from the point to be located to each anchor node, and then establish a virtual anchor according to the distance information of D_actual and D_virtual The position information of the node; if the distance D_virtual between the estimated position to be located and each anchor node is less than the value of D_actual measured by RSSI, the node will be corrected backward according to the difference information, if the estimated position to be located and each anchor node If the distance D_virtual of the node is greater than the value of D_actual measured by RSSI, the node will be corrected forward; if the distance D_virtual between the estimated position of the point to be located and each anchor node is equal to the value of D_actual measured by RSSI, the node does not need to correct the deviation ; Then, according to the position information of the virtual anchor node and the value of D_actual, the ranging algorithm is performed again to judge whether the point to be fixed is in a valid grid area and meet the accuracy requirement. If yes, display the location information and the number of the grid area where it is located. If not, repeat the above-mentioned deviation correction method. Finally the information position and grid area number are displayed.

Beneficial effects: Compared with the prior art, the present invention has the remarkable advantage that the present invention is suitable for positioning and path planning of indoor robots and products requiring positioning, and can well avoid Obstacles and improve positioning accuracy and precise path planning, simplify the algorithm process, provide precise positioning for robots and make path planning more reliable and effective.

Description of drawings

Fig. 1 is the overall step flowchart of the present invention;

Fig. 2 is the flow chart that the present invention sets up different plane test model;

Fig. 3 is a schematic diagram of the rasterization of the indoor plane map of the present invention;

Fig. 4 is a schematic diagram of indoor anchor node layout and distance measurement of the present invention;

Fig. 5 is the effect diagram when the iBeacon layout of the present invention is on the ground, and the measuring point step length is 1m and 0.5m respectively;

Fig. 6 is the effect diagram when iBeacon layout is on the roof in step (a) of the present invention, and the measuring point step length is respectively 1m and 0.5m;

Fig. 7 is a comparison diagram of the iBeacon layout of the present invention on the ground and the roof;

Fig. 8 is a flow chart of the processing calculation of the projection plane of the present invention;

Fig. 9 is a schematic diagram of the distance measurement of the projected height of the anchor node and the distance from the location to be measured in the present invention;

Fig. 10 is a flow chart of the same-plane precision adjustment and correction method of the present invention;

Fig. 11 is a schematic diagram of the deviation correction method in which the virtual node distance of the point to be located is less than the D_actual measured by RSSI in the same plane precision adjustment deviation correction of the present invention;

Fig. 12 is a schematic diagram of the deviation correction method in which the virtual node distance of the point to be located is greater than D_actual measured by RSSI in the same-plane precision adjustment deviation correction of the present invention.

detailed description

In order to describe the technical solution disclosed in the present invention in detail, further elaboration will be made below in conjunction with the accompanying drawings and specific embodiments.

As shown in Figure 1, an indoor positioning method based on different plane anchor node mapping and rasterization correction includes the following three steps:

(a) Establish a different plane test model;

(b) Dimensionality reduction processing of the projection plane obtained by mapping different planes;

(c) Adjust the deviation correction for the accuracy of the same plane.

As shown in Figure 2, in step (a) to establish a different-plane test model, firstly, the different planes are projected onto the plane to obtain the indoor planar map, and then the indoor planar map is uniformly gridded, and then the indoor anchor nodes are arranged, preferably Using iBeacon as the beacon node, the electric field signal strength of the point to be located is collected, converted into distance information, and finally a different plane test model is established. Preferably, the indoor working environment of the home service robot is converted into a grid map, only static obstacles such as walls and partitions in the working environment are considered, and the obstacles are expanded. The state of each grid is divided into three situations: free area, obstacle area, and dynamic area. Among them, the dynamic area is an important area connecting each sub-working area, such as the area where the door is located. By rasterizing the floor plan of the home and numbering the area, the position information can be effectively obtained, which obtains the grid number and establishes the coordinate system of the projection plane, as shown in Figure 3, where Indicates the obstacle area, represents a free area, Indicates a dynamic region. The connectivity of dynamic regions is uncertain, such as the region where the door is located. Then place the iBeacon on the top of the house. Here, iBeacon is preferably used as the anchor node and virtual base station. As shown in Figure 4, the anchor nodes B1, B2, and B3 are arranged on the same level as the indoor roof where the home service robot works, and Vertically project the anchor nodes B1, B2, B3 onto the same plane where the grid map is located to obtain the projection points B1', B2', B3', and the anchor nodes B1, B2, B3 to the anchor node projection points B1', B2' , the vertical height h of B3', collect the RSSI values at different distances from iBeacon, calculate the straight-line distances D1, D2, and D3 between the anchor node and the point P to be located according to the RSSI positioning algorithm, and use Matlab to fit the RSSI and distance The curve is used to simulate the indoor attenuation model suitable for the home environment, and obtain the relationship between RSSI and distance D. As shown in Figure 5, the anchor joints are arranged at the same horizontal position on the indoor ground, and when the measuring point step length is 1m and 0.5m respectively, the effect diagram of the distance from the RSSI to the iBeacon and the measured RSSI signal value is fitted by Matlab . As shown in Figure 6, the anchor joints are arranged at the same horizontal position as the indoor roof, and when the measuring point step length is 1m and 0.5m respectively, the effect of the distance from the RSSI to the iBeacon and the measured RSSI signal value is fitted by Matlab picture.

As shown in Figure 7, the comparison diagram of the effect of placing anchor nodes on the ground and on the roof, it can be seen from the figure that the method of placing iBeacon on the roof is preferred.

As shown in Figure 8, in the dimensionality reduction process of the projection plane obtained by mapping different planes in step (b), the first step is to project the anchor nodes vertically onto the ground, and the second step is to obtain the points to be positioned by testing the model on different planes At this time, the distance D corresponding to the RSSI of P, the third step is to obtain the actual distance D_actual from the distance D through vertical projection, and the fourth step is to obtain the position information of the point P to be located through the ranging algorithm. In the step (b), the distance is measured as shown in Figure 9, the anchor nodes B1, B2 and B3 are arranged indoors, and the anchor nodes B1, B2 and B3 are respectively projected onto the ground to obtain projection points B1', B2 ', B3', the distances D1, D2, and D3 from the anchor node to the point P to be located are obtained through the distance conversion of the established different-plane test model and the RSSI signal value, and then the distances D1, D2, and D3 are respectively projected to the grid On the map, in Figure 9, h represents the height of the anchor node iBeacon from the projection plane, and the distances D1, D2, and D3 are projected onto the same plane to obtain D1', D2', and D3', and D_actual1, D_actual2, and The distance value of D_actual3. The specific algorithm is as follows:

Among them: the height h is the vertical height from the roof to the point to be measured, D_actuaL is the calculated straight-line distance from the projected point P to the anchor node, D is the RSSI value conversion distance of the point P to be positioned, and h is the anchor node distance projection The height of the plane.

Wherein ( _xi , y _i ) is the location information of the anchor node, and (x, y) is the location information of the anchor point.

As shown in FIG. 10 , in the step (c) of adjusting and rectifying the in-plane accuracy, the positioning point information is calculated three times when the positioning point information calculated by the ranging algorithm is not ideal. First, judge whether the positioning point is ideal according to whether the position of the positioning point and the actual position fall in the same grid area. If it is in the same grid area, stop the calculation and display the positioning information and grid area at this time. The specific implementation is as follows:

Calculate the distance D_virtual from the positioning point to each anchor node according to the positioning result. And according to the distance information of D_virtual and D_actual, the position information of the virtual anchor node is established. If the distance D_virtual between the estimated location and each anchor node is smaller than the value of D_actual measured by RSSI, the node will be corrected backward according to the difference information. Otherwise, the node is skewed forward. According to the position information of the virtual anchor node and the value of D_actual, the ranging algorithm is performed again to judge whether the point P to be located is in a valid grid area, and the accuracy requirement is met. If yes, display the location information and the number of the grid area where it is located. If not, continue with deskewing.

Calculate the distance of D_virtual at this time through the information position of the anchor node, the information position of the point P to be located and the distance formula between the two points. And according to the distance between D_virtual and D_actual, the position information of the virtual anchor node is calculated. If the value of the distance D_virtual between the estimated position of the point P to be located and each anchor node is smaller than the value of D_actual measured via RSSI, the realization is as shown in FIG. 11 . The virtual base station is calculated as follows:

and

Otherwise, if the value of the distance D_virtual between the estimated position of the point P to be located and each anchor node is greater than the value of D_actual measured via RSSI, then the realization is shown in Figure 12, and the calculation method of the virtual base station is as follows:

and

Among them, (x', y') is the location information of the virtual anchor node, (x, y) is the initially calculated positioning information, and ( _xi , y _i ) is the information of the initial anchor node.

The acquisition of D_virtual is to make a straight line between the calculated initial position (x, y) of the anchor point P and the anchor nodes B1 (x1, y1), B2 (x2, y2), B3 (x3, y3) respectively. Obtained by the formula, D_virtual1, D_virtual2, D_virtual3, and D_virtual are used during deviation correction to determine whether the deviation correction is inward or outward.

Based on the position information of the virtual node obtained in the previous step, the position information of the positioning node is calculated again. A revised result of the preliminary position estimate is obtained. The same process can be repeated, and the number of iterations is determined according to the grid position and the requirements for positioning accuracy. Through continuous iteration, errors caused by non-line-of-sight and multipath caused by inaccurate RSSI measurement are offset, so as to achieve the purpose of improving positioning accuracy. Finally, the information position and grid area number are displayed to obtain precise positioning with anchor points.

Claims (7)

1. a kind of indoor orientation method based on the mapping of different plane anchor node and rasterizing correction, it is characterised in that：Described side Method includes three below step：

(a) different plane test model is set up；

(b) dimension-reduction treatment of projection plane；

(c) coplanar precision adjusts the coordinate for rectifying a deviation and obtaining point to be determined.

2. a kind of indoor orientation method based on the mapping of different plane anchor node and rasterizing correction according to claim 1, It is characterized in that：Different plane test model of setting up described in the step (a) includes environmental map rasterizing and indoor anchor node Location arrangements.

3. a kind of indoor orientation method based on the mapping of different plane anchor node and rasterizing correction according to claim 2, It is characterized in that：Described environmental map rasterizing is to obtaining indoor plane on coplanar by indoor obstructing objects upright projection Map, position size on indoor plane map is then be mapped to barrier and carries out expansion process, then will be described indoor flat Face map uniform lattice obtains grating map, by the grating map according to the mapping of barrier be divided into barrier zone, from By region and dynamic area, then described grating map is numbered in order successively.

4. a kind of indoor orientation method based on the mapping of different plane anchor node and rasterizing correction according to claim 2, It is characterized in that：Described indoor anchor node arrangement is that iBeacon beaconing nodes are arranged on into roof top, is then determined respectively The location of anchor point is apart from the air line distance of each anchor node and the vertical height of each iBeacon beaconing nodes, and two RSSI value during collection anchor point and iBeacon different distances is surveyed more than secondary, by Matlab fit RSSI and anchor point with The distance Curve of anchor node, sets up the indoor attenuation model of the suitable home environment of simulation, obtains RSSI and anchor point and anchor node Distance relation.

5. a kind of indoor orientation method based on the mapping of different plane anchor node and rasterizing correction according to claim 3, It is characterized in that：Described barrier zone is fixed object, including indoor wall and static places object；Described free space For clear area；Described dynamic area includes door and movable object location of projection region.

6. a kind of indoor orientation method based on the mapping of different plane anchor node and rasterizing correction according to claim 1, It is characterized in that：The dimension-reduction treatment of described step (b) projection plane be by indoor attenuation model, by the anchor point of acquisition with The RSSI value of anchor node is converted into range information, while this distance is projected into two bit planes by Pythagorean theorem, is positioned Point, then by location algorithm and range information D_actual, obtains point to be determined apart from the horizontal linear distance of anchor node Estimated location.

7. a kind of indoor orientation method based on the mapping of different plane anchor node and rasterizing correction according to claim 1, It is characterized in that：The coplanar precision adjustment correction of described step (c) is that the estimated location drawn according to step (b) is calculated Then believed to the positional information and D_virtual of each anchor node according to D_actual and D_virtual distance point to be determined Breath sets up the positional information of virtual anchor node；If by estimated location to be positioned with each anchor node apart from D_virtual Less than the value by the RSSI D_actual measured, then node is rectified a deviation backward according to difference information, if by estimation to be positioned Position and each anchor node apart from D_virtual are more than value by the RSSI D_actual measured, then node are rectified a deviation forward； If the estimated location of point to be determined is equal to by the RSSI D_actual's measured with each anchor node apart from D_virtual Value, then node is without being rectified a deviation；Then according to the positional information of virtual anchor node and D_actual value, ranging is carried out again Algorithm, judges whether point to be located is in effective grid region, reaches required precision；If it is, display location information and institute Locate grid region numbering, if it is not, then re-starting above-mentioned method for correcting error, last display information position and grid region are compiled Number.