High-precision indoor positioning method based on ultra wide band

通过在无GPS室内环境中使用Chan-Taylor算法和粒子群算法计算信号到达时间差权重矩阵,解决了室内定位技术中多径效应和非视距传输导致的精度下降问题,实现了高精度和高效的 Indoor positioning.

CN118921627APending Publication Date: 2024-11-08GUANGDONG POLYTECHNIC NORMAL UNIV

Patent Information

Application Number
CN202410847614.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-06-27
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

In complex indoor environments without GPS, indoor positioning technology is difficult to achieve high-precision positioning due to factors such as multipath effect and non-line-of-sight transmission. The existing algorithms rely on initial values, high computational complexity, large time overhead and reduced accuracy. The problem.

Method used

By setting the base station and the label to be located, the time difference measurement value of the signal arrival is collected, and the Chan-Taylor algorithm and the particle swarm algorithm are used to iterate the signal arrival time difference weight matrix to reduce the impact of non-horizon transmission and improve positioning accuracy.

Benefits of technology

High-precision indoor positioning in complex indoor environments is achieved, positioning errors are reduced, algorithm robustness and computing efficiency are improved, and computing resources are reduced.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118921627A_ABST
    Figure CN118921627A_ABST
Patent Text Reader

Abstract

The invention discloses an ultra-wideband-based high-precision indoor positioning method, which comprises the following steps of: randomly placing a label to be positioned in a base station signal coverage area, and collecting a signal time difference of arrival measurement value of the label; initializing a weight matrix, and substituting the matrix into a preset positioning algorithm for initial position estimation; substituting the initial coordinate of the to-be-positioned label into the signal time difference of arrival equation set to calculate an error vector, and then taking an average value; inputting the obtained average value into a particle swarm algorithm to obtain an optimal weight matrix calculation formula parameter; moving the to-be-positioned label, collecting a signal arrival time difference measurement value of the to-be-positioned label in real time, and inputting the signal arrival time difference measurement value into the weight matrix calculation formula of the optimal coefficient to obtain a weight matrix of the current to-be-positioned label; and substituting the weight matrix of the current to-be-positioned tag into a positioning algorithm to obtain a more accurate estimated position of the to-be-positioned tag.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to a high-precision indoor positioning method based on ultra-wideband. Background Art

[0002] Indoor positioning technology has gradually attracted widespread attention from scholars at home and abroad due to the high requirements of automation and intelligence in the industrial field. In complex indoor environments without GPS, it is extremely challenging to achieve accurate indoor positioning due to the influence of multipath effect (MP) and non-line-of-sight (NLOS) transmission and the dynamic changes of the environment.

[0003] Ultra-wideband (UWB) indoor positioning technology stands out among many indoor positioning technologies with its advantages of high-speed data transmission, low transmission power, energy saving and high security. At present, indoor positioning technologies for wireless signals can be summarized into four categories: based on signal arrival time (TOA), based on signal received strength indication (RSSI), based on signal arrival angle (AOA) and based on signal arrival time difference (TDOA). It is worth noting that the signal arrival time difference method is more advantageous in application because it does not require strict clock synchronization.

[0004] According to the required positioning measurement values ​​collected according to the TDOA principle, a nonlinear equation group is formed, and the positioning problem is transformed into a problem of finding the optimal solution to the nonlinear equation group. When solving the nonlinear TDOA equation group, the Taylor sequence method uses the Taylor formula to fit it to a linear underdetermined equation group near the given initial value and ignores the second-order and above components. The weighted least squares method is used to fit as much as possible to obtain the local minimum sum of squares error, so as to correct the initial value of each step, and continuously iterate to continuously approach the correct value. However, this method has the problem of relying on the initial value and not being able to guarantee convergence and high time cost. The Chan algorithm linearly simplifies the nonlinear TDOA equation group, eliminates the square terms of the unknowns, and only retains the first-order terms to obtain a linear equation group with TDOA noise. Due to the existence of measurement noise, the solution cannot satisfy all equations. The weighted least squares method is used to minimize the sum of squares of the estimated errors and obtain the initial solution. Because in the process of estimating the position for the first time, it is assumed that the unknowns are independent. In fact, there is a correlation between the unknowns. In order to obtain more accurate results, the nonlinear constraint formula between the unknowns is used to perform the second estimation. The Chan algorithm has a fast calculation speed and relatively accurate results, but the derivation of the algorithm is based on the assumption that the channel error is small and obeys a zero-mean Gaussian distribution. This assumption is difficult to meet in a real environment, so the accuracy will drop significantly.

[0005] Then, scholars proposed the Chan-Taylor algorithm by combining the advantages of Chan and Taylor sequence algorithms. The result of Chan positioning method is used as the initial value of Taylor iteration, and the positioning result is gradually approached to the true value through Taylor method. Compared with randomly selecting initial value for Taylor iteration, this method improves the probability of Taylor convergence and improves positioning accuracy. However, when the NLOS error is large, Taylor will make the positioning result deviate further from the true value.

[0006] Comparative document 1: Chinese invention patent with publication number CN103925925B discloses "A real-time high-precision position solution method for a multi-point positioning system".

[0007] Comparative Document 2: Chinese invention patent with announcement number CN109302690A discloses "A non-line-of-sight indoor positioning method based on optimized Kalman filtering".

[0008] Comparative Document 1 and Comparative Document 2 use the Kalman filter algorithm to filter the positioning information, and then substitute it into the positioning algorithm to obtain the final coordinate estimate. The computational complexity of Kalman filtering is high, and a large amount of TDOA data is required to approximate the posterior probability distribution, so a large amount of calculation is required. This will not only increase the calculation time, but also occupy a large amount of computing resources, which may have a greater impact on application scenarios with high real-time requirements. However, this method does not require a large amount of TDOA data. In the face of different positioning environments, only the estimated coordinates of 3 tags to be located are needed to calculate the error value, which is substituted into the particle swarm algorithm for iteration to determine the error weight calculation function, thereby determining the TDOA weight matrix of the remaining tags to be located, and substituting it into Chan-Taylor to obtain the final estimated coordinates.

[0009] Comparative Document 3: The Chinese invention patent with announcement number CN110493742A discloses "A method for indoor three-dimensional positioning for ultra-wideband". Comparative Document 3 uses the preliminary estimation results of the chan algorithm and the base station coordinates to establish a square residual function, construct the corresponding error covariance matrix, and iterate the Gauss-Newton algorithm on the estimated value to effectively suppress NLOS error interference.

[0010] Comparative Document 4: The Chinese invention patent with the announcement number CN105101403B discloses "A precise positioning method based on emergency cellular communication network". In order to eliminate the influence of obstacles, Comparative Document 4 uses Chauvin's method to remove the gross error of the position estimation result, so that the estimation accuracy is further improved.

[0011] Comparative Document 5: The Chinese invention patent with announcement number CN112748397B discloses "UWB positioning method based on adaptive BP neural network under non-line-of-sight conditions". Comparative Document 5 uses UWB radio to perform multiple measurements in indoor LOS environment and NLOS environment respectively, and collects TDOA data in LOS environment and different NLOS environments respectively. The K-means++ clustering algorithm based on unsupervised learning is used to divide the observed data into three categories: LOS, NLOSL and NLOSH. Then, the weights and thresholds of the BP neural network are dynamically optimized through the GA algorithm to further improve the efficiency of the algorithm. Finally, the Chan algorithm is used to solve the initial coordinates of the label, and the calculated coordinates and reconstructed data are used to update the label coordinates by the Taylor iterative algorithm.

[0012] Comparative Document 6: The Chinese invention patent with the announcement number CN113015241B discloses "A TDOA positioning method and system". Comparative Document 6 adopts an iterative balancing algorithm based on enhanced chan algorithm, local space search algorithm and historical information, which can maintain high accuracy and low computational overhead when calculating the location information of the target node, and enhance the robustness of the algorithm.

[0013] Comparative Document 7: The Chinese invention patent with the announcement number CN113255872B discloses "Longhorn Beetle Exploration Positioning Method Based on Chan Algorithm". Comparative Document 7 positions the target using the Chan-Taylor algorithm and the Kalman-Chan algorithm, which are respectively set as a positioning result and the initial position of the longhorn beetle, and the position of the longhorn beetle is continuously updated until the optimal solution is found or the iteration ends.

[0014] Comparative Document 8: The Chinese invention patent with announcement number CN112738709B discloses "A particle swarm search indoor positioning method and system based on UWB signals". Comparative Document 8 introduces a particle swarm search algorithm to improve the positioning accuracy. Under the premise that the UWB positioning error has a positive error, the particle outside the TOA ranging circle is difficult to survive to improve the search accuracy and efficiency.

[0015] The present invention uses the collected TDOA of the three tags to be located to calculate the estimated coordinates and substitute them into the error vector to calculate the estimated error, and then substitutes them into the particle swarm algorithm for iteration to obtain the TDOA weight calculation function, and determine the error weight matrix of different positioning tags, so as to offset the influence of environmental errors and improve positioning accuracy. Therefore, the present invention is different from the solutions of comparative documents 3 to 8.

[0016] Comparative Document 9: The Chinese invention patent with the announcement number CN113589225B discloses "A Method and Device for Positioning an Aerial Target". Comparative Document 9 extracts the content carried by the signal sent by the positioning tag from the positioning base station to obtain the optimal base station configuration that meets the positioning target, and obtains the target position in combination with the positioning algorithm.

[0017] Comparative Document 10: The Chinese invention patent with announcement number CN103987063B discloses "A positioning method for eliminating NLOS errors based on multi-point monitoring". Comparative Document 10 groups all monitoring nodes in the current monitoring area, extracts the coordinates of the target estimated by the monitoring nodes and calculates the geometric precision factor, and selects the combination of monitoring nodes that meet the conditions to improve the positioning accuracy.

[0018] Comparative Document 11: The Chinese invention patent with announcement number CN109613477B discloses "A TDOA Positioning and Tracking Method in Complex Environments". Comparative Document 11 processes the collected positioning data, removes base stations with large errors, solves the data with the number of remaining valid base stations greater than or equal to 4, obtains the position estimation result, and continuously changes the main base station for position estimation, and performs average filtering on the results to obtain the final position estimation.

[0019] Comparative Document 12: The Chinese invention patent with the announcement number CN106162555B discloses "An indoor positioning method and system based on multi-point monitoring to eliminate NLOS error positioning method". Comparative Document 12 divides the possible location area of ​​the mobile station by deploying a sufficient number of base stations to completely cover the indoor environment to be tested. The signal loss through the wall in the same area is the same. The combination with the smallest signal loss estimation error is selected to correct and update the distance between the mobile station and the base station, so as to obtain the accurate distance and positioning coordinates between the mobile station and the base station.

[0020] The present invention does not require additional hardware resources, and only improves the indoor positioning algorithm to improve positioning accuracy. Therefore, the present invention is different from the solutions of reference documents 9 to 12.

[0021] Comparative Document 13: The Chinese invention patent with announcement number CN106851821B discloses "A method for indoor three-dimensional positioning based on wireless communication base stations". Comparative Document 13 first performs three-dimensional indoor coarse positioning based on the CHAN algorithm to obtain TOA data, pre-processes the TOA data through machine learning, eliminates multipath errors based on the LS algorithm, and improves positioning accuracy. This patent uses TOA positioning information, which requires precise time synchronization function and has large errors. The present invention uses TDOA positioning information, which does not require a special timestamp for clock synchronization compared to TOA, and improves positioning accuracy. Summary of the invention

[0022] In order to solve the technical problem that the indoor positioning in the prior art is difficult to achieve accurate indoor positioning in a complex indoor environment without GPS due to factors such as multipath effect and non-line-of-sight transmission (NLOS), the present invention provides a high-precision indoor positioning method based on ultra-wideband, which includes the following steps:

[0023] Several base stations are set up in the application area, and tags to be located are randomly placed in the application area;

[0024] Changing the position of the tag to be located, collecting signal arrival time difference measurement values ​​of several groups of tags to be located;

[0025] Set the signal arrival time difference weight matrix to the unit matrix and initialize the parameters of the signal arrival time difference weight matrix calculation formula;

[0026] Substituting the signal arrival time difference weight matrix into a preset Chan-Taylor algorithm for position estimation to obtain the initial coordinate positions of several tags to be located;

[0027] The coefficients of the signal arrival time difference weight matrix calculation formula are calculated by using the initial coordinate positions of several tags to be located, thereby obtaining the TDOA weight matrix calculation formula;

[0028] Collect the signal arrival time difference measurement values ​​of the current tag to be located in real time, input the set of real-time signal arrival time difference measurement values ​​and the optimal signal arrival time difference weight matrix calculation formula coefficients into the signal arrival time difference weight matrix calculation formula, and obtain the signal arrival time difference weight matrix of the current tag to be located;

[0029] The signal arrival time difference weight matrix of the current tag to be located and a corresponding set of real-time signal arrival time difference measurement values ​​are substituted into the Chan-Taylor algorithm for position estimation to obtain the position of the current tag to be located.

[0030] As a preferred solution, a method of setting up several base stations in an application area and randomly placing tags to be located in the application area includes:

[0031] Four base stations are set in the application area, and the four base stations are respectively recorded as the first base station (x1, y1), the second base station (x2, y2), the third base station (x3, y3) and the fourth base station (x4, y4); wherein the first base station (x1, y1) is set as the main base station; the coordinates of the tag to be located are marked as (x, y).

[0032] As a preferred solution,

[0033] The Chan-Taylor algorithm includes the Chan algorithm, and the Chan algorithm includes:

[0034] The initial nonlinear signal arrival time difference equation group is converted into a linear equation group, and then the least square method is used to perform the first position estimation, wherein the expression of the initial nonlinear signal arrival time difference equation group is:

[0035]

[0036] Where: c is the propagation speed of electromagnetic waves, Δτ i,1 is the arrival time difference of the signal from the i-th base station and the first base station to the tag to be located, N is the total number of base stations, r i,1 is the difference in arrival distances from the i-th base station and the first base station to the tag to be located; the expression for the first position estimation is:

[0037] z a =(G a T QG a ) -1 G a T Qh(2)

[0038] in,

[0039] Q is the covariance matrix of the delay error;

[0040] Then, a second position estimation is performed based on the first position estimation to obtain a corresponding position estimation value. The second position estimation expression is:

[0041]

[0042] z′=(G′ T Σ′ -1 G′) -1 G′ T Σ′ -1 h′ (5)

[0043] in:

[0044]

[0045] Σ is the noise covariance matrix.

[0046] As a preferred solution,

[0047] The Chan-Taylor algorithm also includes the Taylor algorithm, and the Taylor algorithm includes:

[0048] The estimated value obtained by continuously iterating and calculating the second position estimation by Taylor series expansion method is expressed as:

[0049]

[0050] in:

[0051]

[0052] Q' represents the covariance matrix of the signal arrival time difference measurement value, x0, y0 are initial values, Δx, Δy are measurement errors, and x=x0+Δx, y=y0+Δy.

[0053] As a preferred solution,

[0054] The weight matrix calculation formula is:

[0055] w i,1 = a*exp(b*Δτ i,1 ) + c*exp(d*Δτ i,1 ),i=2…N (10)

[0056] Where: Δτ i,1 is the arrival time difference of the signal from the i-th base station and the first base station; w i,1 is the weight ratio of its corresponding signal; coefficients a, b, c, d are rational numbers;

[0057] The signal arrival time difference weight matrix is ​​expressed as:

[0058]

[0059] The final position estimation formula of the Chan algorithm with the signal arrival time difference weight matrix is ​​expressed as:

[0060] z′=(G′ T W -1 G′) -1 G′ T W -1 h′ (12)

[0061] The position estimation formula of the Taylor algorithm with the signal arrival time difference weight matrix is ​​expressed as:

[0062]

[0063] If |Δx|+|Δy|>γ, let x′=x0+Δx, y′=y0+Δy to update the initial coordinate value and perform iterative calculation until |Δx|+|Δy|<γ. When γ is sufficiently small, the algorithm converges and the estimated coordinates of the tag to be located are obtained.

[0064] As a preferred solution, the method of calculating the coefficients of the signal arrival time difference weight matrix calculation formula by calculating the initial coordinate positions of several tags to be located, thereby obtaining the signal arrival time difference weight matrix calculation formula includes:

[0065] Substituting the initial coordinate positions of several tags to be located into the error vector calculation formula to calculate the error value vector, obtaining several error vectors, and then taking the average value of the several error vectors;

[0066] The average value of the plurality of error vectors is input into a preset particle swarm algorithm to obtain the optimal signal arrival time difference weight matrix calculation formula parameters.

[0067] As a preferred solution,

[0068] Substituting the initial coordinate positions of several tags to be located into the signal arrival time difference equation group to calculate the error value vector, obtaining several error vectors, and then taking the average value of the several error vectors includes:

[0069] Substitute the initial coordinate positions of the three tags to be located into formula (1) to calculate three error vectors, and then take the average value φ of the three error vectors.

[0070] As a preferred solution,

[0071] The method of inputting the average value of the plurality of error vectors into a preset particle swarm algorithm to obtain the optimal signal arrival time difference weight matrix calculation formula parameters includes:

[0072] Step 1: input the average value φ of the three error vectors into a preset particle swarm algorithm;

[0073] Step 2: Set the value range of the four coefficients of the weight matrix calculation formula and initialize the particle swarm parameters;

[0074] Step 3: Calculate the fitness value φ of each particle. The optimal criterion for the fitness value of each particle must satisfy the minimum average value φ of the error vector, and update the individual optimal value and the global optimal value according to the calculated fitness value of each particle;

[0075] Update the position and velocity of each particle based on the individual optimal value and the global optimal value, expressed as:

[0076]

[0077] Among them, v i represents the speed of the particle, rand() represents two random vectors between 0 and 1, x i represents the current position of the particle, c1, c2 represent the learning factor, pbesti Indicates the individual optimal value, gbest i Represents the global optimal value, and the ω adaptive weight factor ranges from [0,1];

[0078] Step 5: Determine whether the preset iteration stop condition is met, where the iteration stop condition is that the error value is less than a preset threshold; if not, execute step 3;

[0079] Step 6: Output the optimal signal arrival time difference weight calculation formula parameters.

[0080] As a preferred solution,

[0081] The particle swarm parameters include:

[0082] Swarm size S, particle size D, inertia weight w, acceleration coefficients c1, c2, maximum speed V max , maximum iteration time T.

[0083] As a preferred solution,

[0084] Substituting the signal arrival time difference weight matrix of the current tag to be located into the Chan-Taylor algorithm to perform position estimation, the method for obtaining the position of the current tag to be located also includes:

[0085] The root mean square error is used as an indicator to evaluate positioning accuracy. The root mean square error is defined as follows:

[0086]

[0087] in is the coordinate position estimated by the algorithm, and x, y are the actual coordinate positions.

[0088] Compared with the prior art, the present invention has the following beneficial effects:

[0089] The present invention substitutes the estimated position errors of several groups of tags to be located into the particle swarm optimization algorithm to iteratively determine the optimal coefficients of the TDOA weight matrix calculation formula, thereby calculating the weights of positioning information with different accuracies, reducing the impact of non-line-of-sight transmission on the positioning results and improving the positioning accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] Figure 1 A flow chart of a high-precision indoor positioning method based on ultra-wideband provided in this embodiment;

[0091] Figure 2 The signal arrival time difference positioning principle diagram provided for this embodiment;

[0092] Figure 3A flow chart of the Chan-Taylor algorithm with a signal arrival time difference weight matrix provided in this embodiment;

[0093] Figure 4 The particle swarm algorithm flow chart provided for this embodiment;

[0094] Figure 5 The experimental environment diagram provided for this embodiment;

[0095] Figure 6 A simulation result diagram of a base station regularly placed environment with few obstacles provided in this embodiment;

[0096] Figure 7 A simulation result diagram of a base station regularly placed in a multi-obstacle environment provided in this embodiment;

[0097] Figure 8 A simulation result diagram of an environment with irregular placement of base stations and few obstacles provided in this embodiment;

[0098] Fig. 9 This is a simulation result diagram of an environment with irregular placement of base stations and multiple obstacles provided in this embodiment. DETAILED DESCRIPTION

[0099] The accompanying drawings are only used for illustrative purposes and are not to be construed as limiting the present invention;

[0100] It should be clear that the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the embodiments of the present application.

[0101] The terms used in the embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to limit the embodiments of the present application. The singular forms of "a", "said" and "the" used in the embodiments of the present application and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more associated listed items.

[0102] When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the attached claims. In the description of the present application, it should be understood that the terms "first", "second", "third", etc. are only used to distinguish similar objects, and do not have to be used to describe a specific order or sequence, nor can they be understood as indicating or implying relative importance. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to the specific circumstances.

[0103] In addition, in the description of this application, unless otherwise specified, "multiple" means two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0104] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0105] Example 1

[0106] Please refer to Figure 1 as well as Figure 2 This embodiment provides a high-precision indoor positioning method based on ultra-wideband, including the following steps:

[0107] Several base stations are set up in the application area, and tags to be located are randomly placed in the application area;

[0108] In a specific embodiment, a method of setting up a plurality of base stations in an application area and randomly placing tags to be located in the application area includes:

[0109] Four base stations are set in the application area, and the four base stations are respectively recorded as the first base station (x1, y1), the second base station (x2, y2), the third base station (x3, y3) and the fourth base station (x4, y4); wherein the first base station (x1, y1) is set as the main base station; the coordinates of the tag to be located are marked as (x, y).

[0110] Changing the position of the tag to be located, collecting signal arrival time difference measurement values ​​of several groups of tags to be located;

[0111] Set the signal arrival time difference measurement value weight matrix to the unit matrix and initialize the parameters of the signal arrival time difference measurement value weight matrix calculation formula;

[0112] Please refer to Figure 3 , substituting the signal arrival time difference measurement value weight matrix into the preset Chan-Taylor algorithm for position estimation to obtain the initial coordinate positions of several tags to be located;

[0113] In a specific embodiment, the Chan-Taylor algorithm includes a Chan algorithm, and the Chan algorithm includes:

[0114] The initial nonlinear signal arrival time difference measurement value equation group is converted into a linear equation group, and then the least square method is used to perform the first position estimation, wherein the expression of the initial nonlinear signal arrival time difference equation group is:

[0115] Where: c is the propagation speed of electromagnetic waves, Δτ i,1 is the arrival time difference of the signal from the i-th base station and the first base station to the tag to be located, N is the total number of base stations, r i,1 is the difference in arrival distances from the i-th base station and the first base station to the tag to be located; the expression for the first position estimation is:

[0116] z a =(G a T QG a ) -1 G a T Qh(2)

[0117] in,

[0118] Q is the covariance matrix of the delay error;

[0119] Then, a second position estimation is performed based on the first position estimation to obtain a corresponding position estimation value. The second position estimation expression is:

[0120]

[0121] z′=(G ′T Σ ′-1 G′) -1 G ′T Σ ′-1 h′(5)

[0122] in:

[0123]

[0124] Σ is the noise covariance matrix.

[0125] In a specific embodiment, the Chan-Taylor algorithm also includes a Taylor algorithm, and the Taylor algorithm includes:

[0126] The position estimation value obtained by the second position estimation is calculated by continuously iteratively using the Taylor series expansion method, and the expression is:

[0127]

[0128] in:

[0129]

[0130] Q' represents the covariance matrix of the signal arrival time difference measurement value, x0, y0 are initial values, Δx, Δy are measurement errors, and x=x0+Δx, y=y0+Δy.

[0131] In a specific embodiment, the weight matrix calculation formula is:

[0132] w i,1 =a*exp(b*Δτ i,1 )+c*exp(d*Δτ i,1 ),i=2…N (10)

[0133] Where: Δτ i,1 is the arrival time difference of the signal from the i-th base station and the first base station; w i,1 is the weight ratio of its corresponding signal; coefficients a, b, c, d are rational numbers;

[0134] The signal arrival time difference measurement value weight matrix is ​​expressed as:

[0135]

[0136] The final position estimation formula of the Chan algorithm with the signal arrival time difference measurement value weight matrix is ​​expressed as:

[0137] z′=(G ′T W -1 G′) -1 G ′T W -1 h′(12)

[0138] The position estimation formula of the Taylor algorithm with the signal arrival time difference measurement value weight matrix is ​​expressed as:

[0139]

[0140] If |Δx|+|Δy|>γ, let x′=x0+Δx, y′=y0+Δy to update the coordinate values ​​and perform iterative calculation until |Δx|+|Δy|<γ. When γ is sufficiently small, the algorithm converges and the estimated coordinates of the tag to be located are obtained.

[0141] Substituting the initial coordinate positions of several tags to be located into the signal arrival time difference measurement value equation group to calculate the error value vector, and then taking the average value of the several error vectors;

[0142] In a specific embodiment, a method of substituting the initial coordinate positions of several tags to be located into the signal arrival time difference measurement value equation group to calculate the error value vector and obtain several error vectors includes:

[0143] Substitute the initial coordinate positions of the three tags to be located into formula (1) to calculate the error value vector; then take the average value φ of the three error vectors.

[0144] Please refer to Figure 4 , inputting the average value of the several error vectors into a preset particle swarm algorithm to obtain the optimal signal arrival time difference measurement value weight matrix calculation formula parameters;

[0145] In a specific embodiment, the average value of the plurality of error vectors is input into a preset particle swarm algorithm to obtain the optimal signal arrival time difference measurement value weight matrix calculation formula parameter method, including:

[0146] Step 1: input the average value φ of the three error vectors into a preset particle swarm algorithm;

[0147] Step 2: Set the value range of the four coefficients of the weight matrix calculation formula and initialize the particle swarm parameters;

[0148] Step 3: Calculate the fitness value φ of each particle. The optimal criterion for the fitness value of each particle must satisfy the minimum average value φ of the error vector, and update the individual optimal value and the global optimal value according to the calculated fitness value of each particle;

[0149] Update the position and velocity of each particle based on the individual optimal value and the global optimal value, expressed as:

[0150]

[0151] Among them, v i represents the speed of the particle, rand() represents two random vectors between 0 and 1, x i represents the current position of the particle, c1, c2 represent the learning factor, pbest i represents the individual optimal value, gbest iRepresents the global optimal value, and the ω adaptive weight factor ranges from [0,1];

[0152] Step 5: Determine whether the preset iteration stop condition is met, where the iteration stop condition is that the error value is less than a preset threshold; if not, execute step 3;

[0153] Step 6: Output the optimal signal arrival time difference measurement value weight calculation formula parameters.

[0154] Collecting signal arrival time difference measurement values ​​of the current tag to be located to obtain a set of real-time signal arrival time difference measurement values, inputting the set of real-time signal arrival time difference measurement values ​​and the optimal signal arrival time difference measurement value weight matrix calculation formula coefficients into the signal arrival time difference measurement value weight matrix calculation formula to obtain the signal arrival time difference measurement value weight matrix of the current tag to be located;

[0155] Substituting the signal arrival time difference measurement value weight matrix of the current tag to be located into the Chan-Taylor algorithm to perform position estimation to obtain the position of the current tag to be located;

[0156] In a specific embodiment, the method of substituting the signal arrival time difference measurement value weight matrix of the current tag to be located into the Chan-Taylor algorithm for position estimation to obtain the position of the current tag to be located also includes:

[0157] The root mean square error is used as an indicator to evaluate positioning accuracy. The root mean square error is defined as follows:

[0158]

[0159] in is the coordinate position estimated by the algorithm, and x, y are the actual coordinate positions.

[0160] Example 2

[0161] This example verifies and analyzes the method, more specifically:

[0162] This embodiment conducts positioning experiments in four different indoor environments. In each environment, the coordinate positions of 15 tags to be positioned are estimated, and the root mean square error of the estimated position is calculated. Finally, the positioning performance of the high-precision indoor positioning algorithm and the Chan-Taylor algorithm is compared using the error cumulative distribution function (CDF).

[0163] To test the performance of the high-precision indoor positioning algorithm in complex indoor environments, conduct indoor positioning experiments in different environments. Please refer to Figure 5 , Figure 5 For the experimental environment.

[0164] (1) Base stations are placed regularly in an environment with few obstacles

[0165] Four UWB base stations are placed around an open space, with base station A located at the origin of the coordinate axis. The remaining three base stations are: base station B: (5,0), base station C: (0,5), and base station D: (5,5). 15 tags to be located are randomly placed in the center of the venue, and the signal arrival time difference data of the tags is collected. The collected data set is uploaded to the simulation platform MATLAB. The simulation results are shown in Figure 2. Figure 6 As shown in the figure, the positioning error of the high-precision indoor positioning algorithm is 24 cm, while for the Chan-Taylor algorithm, only 75% of the tags to be positioned have a positioning error less than 24 cm.

[0166] (2) Base stations are regularly placed in an environment with many obstacles

[0167] Four UWB base stations are placed around the obstacle site, with base station A located at the origin of the coordinate axis. The remaining three base stations are: base station B: (5,0), base station C: (0,5), and base station D: (5,5). 15 tags to be located are randomly placed in the center of the site, and the signal arrival time difference data of the tags is collected. The collected data set is uploaded to the simulation platform MATLAB. The simulation results are shown in Figure 1. Figure 7 As shown, for the high-precision indoor positioning algorithm, the positioning error is 34 cm, while for the Chan-Taylor algorithm, only 65% ​​of the tags to be positioned have a positioning error less than 34 cm.

[0168] (3) Base stations are irregularly placed in an environment with few obstacles

[0169] Four UWB base stations are placed around the obstacle site, with base station A located at the origin of the coordinate axis. The remaining three base stations are: base station B: (5,0), base station C: (-1,6), and base station D: (5,6). 15 tags to be located are randomly placed in the center of the site, and the signal arrival time difference data of the tags is collected. The collected data set is uploaded to the simulation platform MATLAB. The simulation results are shown in Figure 1. Figure 8 As shown in Figure 2, the positioning error of the high-precision indoor positioning algorithm is 22 cm, while for the Chan-Taylor algorithm, only 30% of the tags to be positioned have a positioning error less than 22 cm.

[0170] (4) Base stations are irregularly placed in an environment with many obstacles

[0171] Four UWB base stations are placed around the obstacle site, with base station A located at the origin of the coordinate axis. The remaining three base stations are: base station B: (5,0), base station C: (-1,6), and base station D: (5,6). 15 tags to be located are randomly placed in the center of the site, and the signal arrival time difference data of the tags is collected. The collected data set is uploaded to the simulation platform MATLAB. The simulation results are shown in Figure 1. Fig. 9 As shown, for the high-precision indoor positioning algorithm, the positioning error is 34 cm, while for the Chan-Taylor algorithm, only 68% of the tags to be positioned have a positioning error less than 34 cm.

[0172] The above simulation results show that in different environments, the accuracy of the high-precision indoor positioning algorithm is significantly higher than that of the Chan-Taylor algorithm. In particular, in the positioning experiment in an environment with irregular placement of base stations and few obstacles, the positioning error of the high-precision indoor positioning algorithm is 22cm, while for the Chan-Taylor algorithm, only 30% of the tags to be positioned have a positioning error less than 22cm.

[0173] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. A high-precision indoor positioning method based on ultra-wideband, characterized in that: The following steps are involved: Several base stations are set up in the application area, and tags to be located are randomly placed in the application area; Changing the position of the tag to be located, collecting signal arrival time difference measurement values ​​of several groups of tags to be located; Set the signal arrival time difference weight matrix to the unit matrix and initialize the parameters of the signal arrival time difference weight matrix calculation formula; Substituting the signal arrival time difference weight matrix into a preset Chan-Taylor algorithm for position estimation to obtain the initial coordinate positions of several tags to be located; The coefficients of the signal arrival time difference weight matrix calculation formula are calculated by using the initial coordinate positions of several tags to be located, thereby obtaining the signal arrival time difference weight matrix calculation formula; Collect the signal arrival time difference measurement values ​​of the current tag to be located in real time, input the set of real-time signal arrival time difference measurement values ​​and the optimal signal arrival time difference weight matrix calculation formula coefficients into the signal arrival time difference weight matrix calculation formula, and obtain the signal arrival time difference weight matrix of the current tag to be located; The signal arrival time difference weight matrix of the current tag to be located and a corresponding set of real-time signal arrival time difference measurement values ​​are substituted into the Chan-Taylor algorithm for position estimation to obtain the position of the current tag to be located.

2. The high-precision indoor positioning method based on ultra-wideband according to claim 1, characterized in that: The method of setting up a plurality of base stations in an application area and randomly placing tags to be located in the application area includes: Four base stations are set in the application area, and the four base stations are respectively recorded as the first base station (x1, y1), the second base station (x2, y2), the third base station (x3, y3) and the fourth base station (x4, y4); wherein the first base station (x1, y1) is set as the main base station; the coordinates of the tag to be located are marked as (x, y).

3. The high-precision indoor positioning method based on ultra-wideband according to claim 2, characterized in that: The Chan-Taylor algorithm includes the Chan algorithm, and the Chan algorithm includes: The initial nonlinear signal arrival time difference equation group is converted into a linear equation group, and then the least square method is used to perform the first position estimation, wherein the expression of the initial nonlinear signal arrival time difference equation group is: Where: c is the propagation speed of electromagnetic waves, Δτ i,1 is the arrival time difference of the signal from the i-th base station and the first base station to the tag to be located, N is the total number of base stations, r i,1 is the difference in arrival distances from the i-th base station and the first base station to the tag to be located; the expression for the first position estimation is: in, Q is the covariance matrix of the delay error; Then, a second position estimation is performed based on the first position estimation to obtain a corresponding position estimation value. The second position estimation expression is: z′=(G′ T In -1 G′) -1 G′ T In -1 h′ (5) in: Σ is the noise covariance matrix.

4. The high-precision indoor positioning method based on ultra-wideband according to claim 3, characterized in that: The Chan-Taylor algorithm also includes the Taylor algorithm, and the Taylor algorithm includes: The estimated value obtained by continuously iterating and calculating the second position estimation by Taylor series expansion method is expressed as: in: Q' represents the covariance matrix of the signal arrival time difference measurement value, x0, y0 are initial values, Δx, Δy are measurement errors, and x=x0+Δx, y=y0+Δy.

5. The high-precision indoor positioning method based on ultra-wideband according to claim 4, characterized in that: The weight matrix calculation formula is: w i,1 = a*exp(b*Δτ i,1 ) + c*exp(d*Δτ i,1 ),i=2…N (10) Where: Δτ i,1 is the arrival time difference of the signal from the i-th base station and the first base station; w i,1 is the weight ratio of its corresponding signal; coefficients a, b, c, d are rational numbers; The signal arrival time difference weight matrix is ​​expressed as: The final position estimation formula of the Chan algorithm with the signal arrival time difference weight matrix is ​​expressed as: z′=(G ′T W -1 G′) -1 G ′T W -1 h′(12) The position estimation formula of the Taylor algorithm with the signal arrival time difference weight matrix is ​​expressed as: If |Δx|+|Δy|>γ, let x′=x0+Δx, y′=y0+Δy to update the initial coordinate value and perform iterative calculation until |Δx|+|Δy|<γ. When γ is sufficiently small, the algorithm converges and the estimated coordinates of the tag to be located are obtained.

6. The high-precision indoor positioning method based on ultra-wideband according to claim 5, characterized in that: The method of calculating the coefficients of the signal arrival time difference weight matrix calculation formula by calculating the initial coordinate positions of several tags to be located, thereby obtaining the signal arrival time difference weight matrix calculation formula includes: Substituting the initial coordinate positions of several tags to be located into the error vector calculation formula to calculate the error value vector, obtaining several error vectors, and then taking the average value of the several error vectors; The average value of the plurality of error vectors is input into a preset particle swarm algorithm to obtain the optimal signal arrival time difference weight matrix calculation formula parameters.

7. The high-precision indoor positioning method based on ultra-wideband according to claim 6, characterized in that: Substituting the initial coordinate positions of several tags to be located into the signal arrival time difference equation group to calculate the error value vector, obtaining several error vectors, and then taking the average value of the several error vectors includes: Substitute the initial coordinate positions of the three tags to be located into formula (1) to calculate three error vectors, and then take the average value φ of the three error vectors.

8. The high-precision indoor positioning method based on ultra-wideband according to claim 7, characterized in that: The method of inputting the average value of the plurality of error vectors into a preset particle swarm algorithm to obtain the optimal signal arrival time difference weight matrix calculation formula parameters includes: Step 1: input the average value φ of the three error vectors into a preset particle swarm algorithm; Step 2: Set the value range of the four coefficients of the weight matrix calculation formula and initialize the particle swarm parameters; Step 3: Calculate the fitness value φ of each particle. The optimal criterion for the fitness value of each particle must satisfy the minimum average value φ of the error vector, and update the individual optimal value and the global optimal value according to the calculated fitness value of each particle; Update the position and velocity of each particle based on the individual optimal value and the global optimal value, expressed as: Among them, v i represents the speed of the particle, rand() represents two random vectors between 0 and 1, x i represents the current position of the particle, c1, c2 represent the learning factor, pbest i represents the individual optimal value, gbest i Represents the global optimal value, and the ω adaptive weight factor ranges from [0,1]; Step 5: Determine whether the preset iteration stop condition is met, where the iteration stop condition is that the error value is less than a preset threshold; if not, execute step 3; Step 6: Output the optimal signal arrival time difference weight calculation formula parameters.

9. The high-precision indoor positioning method based on ultra-wideband according to claim 8, characterized in that: The particle swarm parameters include: Swarm size S, particle size D, inertia weight w, acceleration coefficients c1, c2, maximum speed V max , maximum iteration time T.

10. The high-precision indoor positioning method based on ultra-wideband according to claim 9, characterized in that: Substituting the signal arrival time difference weight matrix of the current tag to be located into the Chan-Taylor algorithm to perform position estimation, the method for obtaining the position of the current tag to be located also includes: The root mean square error is used as an indicator to evaluate positioning accuracy. The root mean square error is defined as follows: in is the coordinate position estimated by the algorithm, and x, y are the actual coordinate positions.

Citation Information

Patent Citations

  • A real-time high-precision position calculation method for multi-point positioning systems

    CN103925925B

  • A NLOS error elimination positioning method based on multi-point monitoring

    CN103987063B

  • A precise positioning method based on emergency cellular communication networks

    CN105101403B

  • Indoor positioning methods and systems

    CN106162555B

  • An Indoor 3D Positioning Method Based on Wireless Communication Base Stations

    CN106851821B

Cited By

  • High-precision anti-interference distance measurement method based on UWB

    CN119805429A

  • UWB distance adaptive weighted positioning method for indoor corridor area

    CN119967356A

  • A UWB distance-adaptive weighted localization method for indoor corridor areas

    CN119967356B