TW201126194A - A method of positioning a RFID tag using spatial mesh algorithm - Google Patents

Abstract

A method of positioning a RFID tag using spatial mesh algorithm is disclosed. Radio Frequency Identification (RFID) readers are arranged in a space having a RFID tag therein to be positioned. To position the RFID tag, the space is then divided into a plurality of rectangular spatial boxes. Calculating each distance between a center of rectangular spatial box and each RFID reader is followed to obtain a set of distances for each RFID reader. The differences between the set of distances by calculation for each reader and the distance by measurement are compared and analyzed to find a set of rectangular spatial boxes having minimum differences for each reader. By comparing the set of rectangular spatial boxes having minimum differences for each reader, the position of the RFID tag is determined.

Description

201126194 VI. Description of the invention: [Technical field to which the invention pertains] The present invention is based on - a type of scale _ _ _ ^; Radk) F which _

The Identification label 'specially refers to a kind of algorithm that still locates the location. [Prior Art]

In the 21st century, wireless communication technology is widely used in daily life, which brings us a very large view. In addition to communication, wireless technology can be used. Positioning technology, commonly used positioning technology includes Gps, , infrared, IEEE 802.n, ultrasonic, ultra-wideband, 钿^ and radio frequency identification (RFID; Radio Frequency Identificati〇n). Although Gps can be accurately positioned and low cost, the technology is _ outdoor positioning; CdnD and ultra-wideband are suitable for large fan _ ship miscellaneous; red light shirt interference and pro-county high; loss = 2.11 and 2@Bee positioning effect is not as good; Ultrasonic New Zealand construction costs to 0 RFID wireless rewards tag is a kind of miscellaneous automatic identification (four) system, it is = radio waves to transmit identification data, - group radio frequency knowledge is composed of standard iron and read two ... 'tag tops There are circuits, reading and writing miscellaneous - Duan Lin Reede launch energy = label reading and writing money messages, the standard county is installed on the "cut wafer" early days, the line 'weaving with Lin or Naosaki package. The RFID internal positioning system is proposed by HighT_r and Bie Ting, and the 5H research and development SpotON positioning system verifies the indoor positioning. In the method of Sp identification, the location of the unknown object has not been controlled by the central control system. The process, but the other sensing points of the same hardware specifications, completed with the 201126194 $ scattered square wire, _ scattered face view Dick signal strength (RS clear material collection and return, and finally with positioning; The predicted position of the unknown object is different. ^' RFID days are tilted over the room rainbow construction, using single-_ antenna for positioning, the obtained ^2;»= ϊΓΓΓ"TM antenna' can locate the target location mine = The parent of the surface is a circle of arcs. Adding a third day to the wireless sensing network will have two intersections with this arc. These two points represent the location of the call # _ possible position In order to understand and understand, it is generally necessary to have 4 antennas. The concept requires at least three signal transmission towers to be launched and the coordinates of the signal transmission tower (4), Y=G), (χ==2 (4) , Υ,, the three nodes cover the range ri, the intersection of the points is Calculate the signal at the unknown location Wei _ turn M secret atefati (m read the pepper in the four _ _ case 97_ proposed algorithm is the concept of space position correction, first measure each anti-PID; start c ' The positioning space is generated - the distance of the starting point label is followed by the distance. Then the line from the starting point to the end of each RFID antenna

Root mean square error of Sa F ° antenna and target tag (Root Mean

Sq job E bribe, RMSE). If the root mean square error is small ^ iteratively modified, the iterative correction, A, military field, 仏 、, starting coordinates and adjustment rate calculation. The surface is initialized with the initial coordinate value local gradient, and then the root mean square error of all antennas and target labels is calculated, otherwise the iterative correction is continued. [S1 5 201126194] SUMMARY OF THE INVENTION It is an object of the present invention to provide an RPID tag positioning algorithm that can be applied to two-dimensional or three-dimensional space. The invention relates to a method for positioning radio frequency identification (RFID) tags, comprising at least the following steps: (a) providing three antennas and respectively measuring the measured distances Sk (k=1_3) of the antennas to a target tag; (b) cutting a positioning space containing the target target into n grids; (c) calculating a linear distance Sjk (Hn ' k=l-3) of the three antennas from the center points of the n grids, The actual distance from the 3n antennas to the center of the grid can be obtained; (d) the calculation error %' is the difference between the measurement distance Sk of the antennas and the linear distance of the grid centers. The error, ie sk-Sjk (j=ln, k=l_3); (e) Comparing the error values calculated for each antenna, three sets of meshes with the smallest error value can be obtained, and there are multiple in each group. The error values of the grid are the same; and (7) aligning the grids with the smallest error values in different groups to select the grid from which the intersections are the positions of the target labels. The invention also provides another method for radio frequency identification (RPJD) tag positioning, which comprises at least the following steps: (a) providing three antennas' and respectively measuring the measured distances Sk of the antennas to a target tag (k=1_3) (b) cutting a three-axis pair of a positioning space into an N-block grid; (c) calculating a straight-line distance Sjk (HN, k=l-3) of the three antennas from the center points of the n grids, respectively. Obtaining 3N antennas to the distance between the two of the grids; (d) calculating the error ^, which is the error between the measured distance Sk of the antennas and the linear distance between the centers of the grids, ie 々 = VSjk G=i_N 'b (9); (e) Comparing the error values calculated by each antenna, three sets of meshes with the smallest error value can be obtained, and the error values of multiple nets in each group are also the minimum. (f) Aligning the grids with the smallest error values in different groups from which the grids intersecting each other are selected, and the grid is further fine-cut into blocks f 6 201126194 Grid 'Repeat (c) ~ (f) The step 'determines the position of the target tag until the number of iterations or / and the average error of the selected grid position is less than - set. A further understanding of the advantages and spirit of the invention will be apparent from [Embodiment] - Wireless radio frequency identification (RFID) reading and writing ^ includes an antenna, which can be used to read the radio receiving signal strength of the RFID tag (hereinafter referred to as Rssi), by which the distance can be derived but the position of the RHD target tag is unknown. Therefore, as in the prior art, the 'money wheel set, at least three antennas (but there are still two possible positions)' is understood to be understood, and the embodiment of the present invention is used to obtain three different types of positioning. The unique location of the target. The method of the present invention is called a spatial positioning algorithm 2 S(SpA2 〇), which is to search for a positioning space by using the contiguous-searching space, and to cut the positioning space into a plurality of grids, and calculate the error of each grid sequentially, the smallest in the county. Set for the target age. The embodiment of the present invention can refer to the flowchart shown in FIG. 2. ^ Step 100: Deploy a plurality of reference labels and 4 antennas in a target-containing "signal positioning Lai" for the calculation process to cover various shouts, that is, the antennas are distributed in a uniform manner as much as possible in different locations within the positioning space. The RSSI 'reduced error is obtained in different directions. In addition, the antenna price is higher, and the uniform distribution can reduce the number of antenna arrangements'. Reduce the cost of wireless sensing network construction. Step 105: Since the positions of all reference labels are known, The antenna is also known. Therefore, the radio wave received signal strength (RSSI) values of the reference antennas of the antennas and known locations are measured one by one, and then based on the RSSI of the antennas and the distances of the reference labels. RSSI value-distance relationship diagram, that is, the wireless signal attenuation curve. Because the coffee signal is affected by environmental factors, RSSI and distance (5) 7 201126194 is changed with the environmental cake, and it is necessary to request the relationship between RSSI and the specific environment. Measuring the measured distance between the RFID antenna and the target tag by using this curve

Sk(k=l-4). In one embodiment, a specific relationship of Rssi to distance is plotted with nine reference labels, as shown in FIG. Step 110: Divide the positioning space into three squares and cut into a square grid. The positioning space may also be converted into n square cubes. If the volume is: the predetermined average error preset by the user. The silk filament determines η^小' where the predetermined average error tolerance can be used as a reference using the root mean square error value; Step 115. Calculate the linear distance Sjk of the four antennas from the center point of the η grids respectively (j=ln,k =l-4) 'After obtaining 4n antennas to the center of the grid, step 120 · Calculating the error, the error is the measured distance Sk of the antenna T for the target label and the center of the grid The linear distance is known as the error, that is, the bow Sjk 〇=1·η, k=l-4), Step 125: For each antenna to calculate the error value, the four sets of error values can be obtained. There may be a plurality of grids in the group having the same error value; and step 130: comparing the grids with the smallest error values in the different groups, and selecting the grid of the intersections as the position of the target label. ^ Alternatively - the alternative method is to calculate the average error bow, and the grid with the smallest error is selected as the target miscellaneous. In the embodiment of the present invention, it is - root mean square error (RMSE), and the formula is as follows: m Σ K jk=\ i S1 8 201126194 Line quantity: two = straight line distance. In fact, the root mean square error == is the grid of the intersections of the ‘, , and ^ methods after the comparison. 21f invention advances - step to improve this embodiment 'derived spatial positioning algorithm = 8^2]) 'Previous spA2 () calculates the target target position in an exhaustive manner, which will take a lot of time when it is large. The way of the grid = time 'but the results of the large grid calculation will reduce the accuracy, in addition to the exhaustion of the cost of the time to calculate the irrelevant area, reducing the search function. Therefore, the SPA 2.1 calculus improves the SPA 2. The way to search for all possible locations of the space by g method is different from that of 2. 〇, SPA2.1 first divides the overall space into N blocks 'N is about 4~8, calculate After the error of the center point of each block, the block with the smallest selection error is further divided into the block, wherein Μ is not equal to N. In the embodiment of the present invention, the whole space is equally divided into eight blocks, and the process is as shown in FIG. Show. Step 200: Using the wireless signal attenuation curve drawn in step 11 to find the measurement distance Sk (k=M) of the position of the k-th RFID antenna and the target tag; Step 205: first cut the space three-axis pair into eight networks. Step 210: Calculate the distance between the four antennas and the center point of the eight grids by Sjk〇=l-8 'k=l-4) respectively to obtain the actual distance between the 32 antennas and the center of the grid; Step 215: Calculating the error 々, which is the error between the measured distance sk of the antennas and the linear distance between the centers of the grids, ie, $ (j=l-8, k=l-4); 7 " Jk Step 220: Comparing the error values calculated by each antenna, obtaining four grids with the smallest error value, and the error values of the plurality of grids in each group may be the same; m Step 225: Comparison In the different groups, the grid with the smallest error value, from which 9 201126194 selects the grid of mutual intersections, returns to the step block grid, repeats the above 12Q = and then finely cuts the grid into eight grids _ = heart 225 'Langdang iteration number, she selected the network step (four) - _, when the target is marked: the position of the overlap, where the root mean square error The number of iterations of the same formula is calculated as the number of repetitions of steps 205 to 225.

The most appropriate way to choose the difference is to directly calculate the root mean square county, the root mean square error step is moved, for the grid, then fine-domain-like mesh, square ^ #ϋ5~220 'When iteration number or / and the selected network When the center position of the grid j is taken from the net & ST slave value, the target value of the money (4) is the selected & Γ疋 iteration number to ensure that the positioning is completed within a certain time, and the average error is small;疋 & bit accuracy ' can be selected by the user as needed. In order to verify the feasibility of the above spatial positioning algorithms 20 and 21 (SPA 20 and SPA 2.1) 'The present invention uses the actual case simulation calculus _, the positioning space used is 926 cm long and 535 cm high 211 (10), the actual position of the target standard drug For (694 cm, 400 cm, 75 em). When using SPA 2.0 calculus, first, use the exhaustive method to cut the space into a grid. In this experiment, each grid is cut into cubes of 3〇cm, width 3〇(10), height 3〇(10), and the positioning space is long. 31 grids, width 18 grids, height 8 grids, total 4464, the minimum error of the cube ' § 丨 Μ Μ · · 若 若 若 若 若 若 若 若 若 若 若 若 若 若 若 若 若 若 若 若 若 若 若 若 若 若 网格 若 若 若 若 若The width is (10), the cube is 100 cm high. The positioning space is 1 inch long, 6 squares wide, and 3 square meters high. There are 180 vertical prefixes. The most secret difference after calculation is I.5. Use SPA 2.1 to first cut the positioning space into eight blocks, and use each block to make the heart point different error value. In this eight blocks, select a block that is closest to the target location in 201126194 and then cut eight blocks. The difference is less than a certain value - the value stops. This algorithm can improve the wide-ranging catch of SPAZ0. Because spA2. only needs to cut eight blocks at a time, the convergence speed is faster. The invention uses spA21 to position the eight blocks of the positioning space to the eight centroids of the workpiece, as shown in Fig. 5, and finds the stop after the error calculation: the area error rate is the smallest, as shown in Fig. 6, if the error is satisfied If the stop bar does not satisfy the stop condition, the seventh block is further processed. |J into eight grids until the stop condition is met, and the spA 2 convergence curve is shown in Figure 7. The above is only the preferred embodiment of the present invention. = Others that do not depart from the present invention == Equivalent changes or modifications shall be included in the scope of the following claims. [Simple description of the schema] Schematic diagram of three-dimensional spatial positioning; ^Based on the spatial positioning algorithm of the present invention, PA20) algorithm flow Figure 3 shows the antenna wireless signal attenuation curve. Figure 4 is a perspective view of a spatial orientation in accordance with the present invention. ~ 2 (1) (SPAZ1) algorithm flow :: Show: positioning algorithm 2J grid coordinates. Figure 颂妓 2.1 Errors. Surplus

One of the positioning spaces, the x-angle clamp algorithm, the Z1 convergence trajectory, and the initial position is [Main component symbol description J VII, patent application scope · IS]

Claims (1)

201126194 The block is cut again. The block is less than a certain value. This algorithm can improve the wide-ranging catch of SPAZ0. Because spA2. only needs to cut eight blocks at a time, the convergence speed is faster. The invention uses spA21 to position the eight blocks of the positioning space to the eight centroids of the workpiece, as shown in Fig. 5, and finds the stop after the error calculation: the area error rate is the smallest, as shown in Fig. 6, if the error is satisfied If the stop bar does not satisfy the stop condition, the seventh block is further processed. |J into eight grids until the stop condition is met, and the spA 2 convergence curve is shown in Figure 7. The above is only the preferred embodiment of the present invention. = Others that do not depart from the present invention == Equivalent changes or modifications shall be included in the scope of the following claims. [Simple description of the schema] Schematic diagram of three-dimensional spatial positioning; ^Based on the spatial positioning algorithm of the present invention, PA20) algorithm flow Figure 3 shows the antenna wireless signal attenuation curve. Figure 4 is a perspective view of a spatial orientation in accordance with the present invention. ~ 2 (1) (SPAZ1) algorithm flow :: Show: positioning algorithm 2J grid coordinates. Figure 颂妓 2.1 Errors. One of the locating space x-angle 演 algorithm Z1 convergence trajectory, the initial position is in [main component symbol description J VII, patent application scope · IS] 201126194 1. A method for radio frequency identification (RFID) tag positioning, at least The method includes the following steps: (a) providing three antennas and respectively measuring the measured distances sk (k=l-3) of the antennas to a target label; (b) cutting a positioning space containing the target label into ^c grid; (c) Calculate the straight distance SjicCM η ' k-1-3) of the three antennas from the center point of n grids respectively to obtain the actual distance of 3n antennas to the center of the grid; (d) 言=calculus error%' These errors are respectively made by the antennas to the target tag; the error between the ship Sk and the linear distances of the mesh towels ^, ie ejk~~ Sk~ Sjk (j^ln jk=l-3); (e) Comparing the error values of the parent antennas, the minimum error value of the three sets of error values is obtained, and the error of each of the complex _ lattices is the smallest; (f) Aligning the different sets of the masks with the smallest error value, and selecting the grid from which the intersections intersect is the position of the target label. 2. The method of claim 2, wherein the method further comprises the steps of: measuring the received signal strength of the antenna and the reference target of the known positions by using the label of her label; (RSSI) value, according to the antenna value of the antennas (4) = test label distance to make RSSI value-distance relationship map; and according to the RSSI value of the target tag measured by the antenna, the line corresponding to the value - distance The diagram gets the target _' giant away. 201126194 3. According to the method described in claim 2, the antennas are distributed in a uniform sentence _ granules, and are set in the same position to obtain the 参考海 reference label or the RSSI of the target label. value. 4. If you apply for a patent scope! The method according to the item, wherein in the step (4), when the positioning space is cut into n grids, the three partitions of the positioning space may be equally divided into n square cubes or cut into n square cube grids. If there is a succession space, it is regarded as a grid. ^^申/ΠΓ The method described in item 1 of the scope, the position of the target label in step w is also the grid with the smallest error after the same scale error. In the method described in the item i of the patent scope, after the grid 1 with the smallest error is set as the position of the marshalling label, the following steps are further included: (g) m meshes for the selected mesh and fine mesh, where m No: two:: complex: (cHg), until the selected grid position in the space is a difference of £j or / and the iteration number is less than a set value and then stops. == The method described in the difference of 6 items, wherein the step is averaged 8. The predetermined average error = the method according to the method of claim 8 is allowed to be referenced by a root mean square error. 201126194 1〇. As claimed in the method of claim 1, the + is further provided with a 4th day line according to the steps (a) to (f) to further determine the position of the target tag. 11. A method for positioning a radio frequency identification (RFID) tag, comprising at least the following steps: (a) providing three antennas and separately measuring a distance Sk of the antennas to a target tag (k=l-3) (b) Cut the three-axis pair of the positioning space into n-grid; the error of the straight-line distance Sjk, ie 々=sk-Sjk (j=lN, (e) for each antenna, calculate ___ and Some grid centers k=l-3); 12. If the method of claim 11 is applied, (c) separately calculate the linear distance of the three antennas from the center point of the N grids, k=l-3) 'Available 4N antennas to the grid The actual distance of the center ' / d ) is the difference of the error 々, the error is the measurement distance of the antennas ~ includes the following steps: wherein the above steps (a) are more than a plurality of reference antennas Within the positioning space; the radio wave reception or strength (RSSI) value of the 201126194 target-specific tags and the reference tags of the known locations are based on the RSSI value of each antenna and the distance of the reference tags, respectively. A clear value_distance relationship diagram is generated; and the RSSI value of the target tag measured according to the 5th antenna is obtained according to the RSSI value-distance relationship graph. 13. The method of claim 12, wherein the antennas are distributed in the location space and are disposed at different locations to obtain the RSSI value of the reference tag or the target tag in different orientations. 14. If the method of claim 11 is applied, the method of claim 11 is equal to N, N=4~8 〇2, and the method of item 11 is further provided - the fourth antenna, %, (a) ( f) Step 'to position the target label more accurately. 16·If applying for the _ u method, the root mean square error. In the method described in item 11 of the encyclopedia, the position in step (7) is also the grid with the smallest error after calculating the error of the WT square root of the large phase. U1 15