CN108307301A - Indoor orientation method based on RSSI rangings and track similitude - Google Patents

Abstract

The invention discloses an indoor positioning method based on RSSI (Received Signal Strength Indication, received signal strength indication) distance measurement and trajectory similarity, firstly, the signal noise is processed through a filtering model, and the RSSI representative value of each group of base stations is calculated; secondly, according to The theoretical model of wireless signal transmission and the relationship between the distance from the point to be positioned to each base station establishes a set of correlation equations, and calculates the initial coordinate estimation of the point to be positioned; finally, the initial coordinate estimation is used to find its similar trajectory, and the one similar to the current trajectory is selected from it. The three similar trajectories with the highest degree are drawn, and the coordinate area of the point to be located is delineated accordingly, and the actual coordinates of the point to be located are calculated by combining the coordinate area with the initial coordinate estimation. The positioning method in the present invention attaches great importance to the correlation between the current trajectory and the historical trajectory, while most classical methods regard each positioning calculation as an independent process. Through the method of the present invention, a correlation model between the current location and the historical location is established, and the current location is determined by using the correlation model.

Description

Indoor Positioning Method Based on RSSI Ranging and Trajectory Similarity

technical field

The field of industry involved in the present invention is the field of indoor positioning, which is suitable for indoor positioning application scenarios where the indoor layout is relatively fixed and the movement trajectory of indoor objects or people is relatively stable; the technical field relates to the Internet of Things technology, in particular to an RSSI-based ranging and Indoor localization methods based on trajectory similarity.

Background technique

The development and gradual maturity of modern information technologies such as wireless communications, computer networks, and smart devices have made applications based on geospatial locations ubiquitous, and people's demand for location information has become increasingly urgent. In recent years, GPS has been widely used to provide high-precision positioning services for outdoor users. However, due to objective factors such as ground buildings and indoor furnishings, indoor users cannot obtain GPS signals. At this time, GPS cannot satisfy indoor users. positioning needs. As an effective supplement to outdoor satellite navigation and positioning technology, indoor positioning technology has certain commonality between them. However, indoor positioning has many interference factors and requires higher positioning accuracy. These characteristics make outdoor positioning methods unable to be directly applied to indoor positioning in complex environments. Indoor positioning method is a key part of indoor positioning technology, and has become a research hotspot in the field of indoor positioning.

The current mainstream indoor positioning technology mainly focuses on improving indoor positioning accuracy, applicability, and reducing use costs. According to the communication means used by the positioning method, the indoor positioning technology can be generally divided into GNSS technology, wireless positioning technology, positioning technology combined with GNSS and wireless positioning, and other positioning technologies. The wireless positioning technology can be divided into infrared positioning, ultrasonic positioning, WiFi positioning, Bluetooth positioning and so on. Among them, Bluetooth-based wireless positioning technology has the advantages of low technical cost, small equipment size, easy integration and popularization, and high positioning accuracy. It can be used in application scenarios such as real-time positioning of objects, cargo tracking, and robot navigation.

The indoor positioning technology based on Bluetooth mainly adopts the polygon positioning method based on RSSI ranging. The basic principle of the method is to calculate the position of the positioning point by measuring the distance between the positioning point and each Bluetooth base station. The positioning device receives the wireless signals sent by each Bluetooth base station, measures the RSSI of each Bluetooth base station, and calculates the straight-line distance to each Bluetooth base station according to the theoretical model of wireless signal transmission. Estimate the position of the positioning device using polygon positioning algorithm and distance constraints. Although this positioning method introduces filtering models such as mean filtering, median filtering, and Gaussian filtering to eliminate part of the signal noise and reduce the impact of wireless signal fluctuations on positioning accuracy to a certain extent, this method relies too much on the measured RSSI values of each Bluetooth base station , in other words, the measurement accuracy of the RSSI value directly determines the accuracy of the positioning algorithm. During the long-term use of the indoor positioning system, a large amount of user historical positioning data is recorded, the association between the current position and the historical positioning data is established, and the potential value of the historical positioning data is tapped to improve the accuracy of indoor positioning.

Contents of the invention

The purpose of the present invention is to overcome the deficiencies of the prior art, propose a kind of indoor positioning method based on RSSI ranging and trajectory similarity, by introducing the association between current trajectory and historical trajectory, reduce the proportion of RSSI in positioning calculation, thereby Indirectly eliminate the influence of part of the signal noise on the positioning accuracy.

The technical solution adopted by the present invention to solve its technical problems is:

An indoor positioning method based on RSSI ranging and trajectory similarity, comprising:

Step 1. Obtain the RSSI data sets of each Bluetooth base station at the location, and group the RSSI data sets according to the signal transmitting base stations corresponding to the collected RSSI data, that is, the RSSI data of the same base station are grouped into one group. Since the degree of wireless signal loss is proportional to the receiving distance, the greater the signal loss, the less RSSI data will be collected, and the lower the positioning accuracy will be. Therefore, it is necessary to analyze the number of RSSI data collected in each group. Generally, 4 to 8 data sets with a large number of RSSI values are selected to participate in the positioning calculation.

Step 2, RSSI data set optimization processing

Due to the influence of objective factors such as Bluetooth base station transmission instability and indoor object occlusion, the received RSSI values are different in different time periods at the same location, which leads to large differences in the values of the data individuals in the same group of RSSI data sets. The number of RSSI data between groups is also different. In order to enhance the reliability of the RSSI value and minimize the error of positioning results caused by signal fluctuations, it is necessary to filter each set of RSSI data sets so that the RSSI values obtained by each set of processing can reflect the positioning more realistically. The distance from the terminal to the base station. The present invention adopts a model processing method combining a Gaussian model and a Kalman model.

Step 2.1, Gaussian model screening

Ranging based on RSSI is generally based on the theoretical model of wireless signal transmission

In the formula, RSSI(d) _dBm represents the signal strength value at the signal receiving place when the distance from the signal generating source is d; RSSI(d ₀ ) _dBm represents the signal strength value at the signal receiving place when the distance from the signal generating source is d ₀ , d ₀ is generally also called the reference distance, and the value is usually 1m in practical applications; α is called the path loss index, which is a parameter related to the influence of the external environment, and its value can be calculated by measurement; β represents external random factors, which are subject to (0, σ ² ) Gaussian distributed random variable.

From the above analysis, it can be seen that the value of the signal strength indicator rssi _j,d collected when the distance from the base station j is d is a random variable that obeys the (0, σ ² ) Gaussian distribution, which is recorded as: rssi _j,d ~ (0, σ ² ), and its probability density function is

Where rssi _{j, d, k} represent the kth signal strength indication value collected when the distance from base station j is d; Indicates the mean value of each group of RSSI data sets collected during the positioning data collection period, and the calculation formula

n represents the number of rssi values selected in each group; σ represents the standard deviation of each group of RSSI data sets, the calculation formula

standard normal distribution function

The basic idea of Gaussian model filtering RSSI is to select the rssi value between the confidence interval corresponding to the high confidence, namely

P(γ ₁ ≤rssi _j,d,k ≤γ ₂ )=1-δ

In practical engineering applications, the high confidence δ is generally taken as 0.6, so the confidence interval [γ ₁ ,γ ₂ ] of the confidence level 1-δ can be deduced, namely

δ indicates a significant level, that is, a small probability value; Indicates that the probability value is When is the critical value in the normal distribution graph, its value can be found in the normal distribution table. Then the base stations are grouped and screened by the Gaussian model, and finally the RSSI data set selected by each group

Step 2.2, Kalman filter smoothing

In order to further improve the stability of RSSI data, it is necessary to use the Kalman filter to smooth the data set screened by the Gaussian model. The received RSSI of the same base station should be the same), based on the a posteriori estimation rssi _j,d,k-1|k-1 of the k-1 state, the k-state a priori estimation rssi _j,d,k|k to be calculated is obtained _-1 , and combined with the observed value rssi _j,d,k of the current k-state RSSI to make a posteriori estimation rssi _j,d,k|k of the k-state RSSI. Kalman filtering consists of two parts: a process model describing the system state and an observation model for observing the system state.

Process Model Formal Representation

rssi _j,d,k|k-1 ＝θrssi _j,d,k-1|k-1 +E ₁ (k)

Where θ represents the system parameters, and the value of θ is 1 according to the system model; E ₁ (k) represents the Gaussian white noise or system error during the transition of RSSI from state k to state k-1, which obeys the (0,Q) Gaussian distribution, Q is the variance of the error.

Formal representation of observation model

rssi′ _j,d,k = ωrssi _j,d,k +E ₂ (k)

where rssi′ _j,d,k represent the adjustment value of the observation model; ω represents the system observation parameter, and the value is 1 according to the system model; E ₂ (k) represents the system observation Gaussian white noise or the system observation error, obeying (0,R ) Gaussian distribution, R is the variance of the error.

Formal expression of Kalman filter

rssi _j,d,k|k ＝rssi _j,d,k|k-1 +K(rssi′ _j,d,k -rssi _j,d,k|k-1 )

Where K represents the Kalman gain, and its calculation formula

Among them, P _k|k-1 represents the error variance of the prior estimation of the k-state process model, and the relationship with the error variance of the posterior estimation of the k-1 state is P _k|k-1 =P _{k-1|k- 1} +Q, parameters R and Q can be obtained through statistical calculation of actual observation data.

For the data set G _j ={rssi _j,d,l |l=1,2,...,m} filtered by the Gaussian model, the data set G′ _j ={rssi _{j,d,l| l} |l＝1,2,…,m}, finally It is used to calculate the distance from the point to be located to the j base station.

Step 3, establish a set of correlation equations from the projection distance equations from the point to be located to the position of each base station

Assuming that the initial coordinates of the point to be located are estimated to be (xi _′ , y _i ′), the obtained in step 2 Substituting into the theoretical model of wireless signal transmission, the distance from point p to base station j can be obtained as d _j . The relative height of the positioned terminal relative to the positioning space in which it is located generally changes little, so the relative height difference between the positioned terminal and the indoor base station can be set as a parameter h ₀ . So far, the equation of projection distance from the point to be positioned to the position of each base station can be obtained by the law of cosines, and the following correlation equations can be established:

The point (x _j , y _j ) represents the coordinates of the base station j.

Step 4, maximum likelihood estimation method to find the approximate solution of the system of equations

Under the ideal condition of no signal loss, there is a unique solution for x _i ′, y _i ′ in the equation system ①. However, the indoor environment is often more complex, there are many indoor furnishings, and there is a certain loss of wireless signals, and there are generally 6 to 8 equations in the equation group ①, so the equation group ① has no solution. For this purpose, the approximate solution of the system of equations can be obtained by using the method of maximum likelihood estimation. In order to further improve the accuracy of the positioning results, a combination method is adopted for the equations in the equation group ①, that is, j-1 equations are selected from j equations to combine into a new equation group, and a total of j new equations are formed Group. The structure of the solution of the new sth system of equations

in

Finally, the average value of all j new equation system solutions is calculated as the initial coordinate estimation of the point to be located:

Step 5, look for similar historical movement trajectories

The indoor spatial layout is often relatively fixed, and indoor objects or people form relatively fixed historical moving trajectories during the long-term movement and positioning indoors, so the current moving trajectories often have great similarities with some historical moving trajectories. Utilizing this association can improve the positioning accuracy of indoor objects or people whose moving trajectories are relatively stable. The specific implementation is as follows:

Step 5.1, find similar trajectories

Find a point set P={(x,y)||xx _i ′|<λ,| _yy i ′|<λ _} that is closer to the initial position estimate (xi ′, y _i ′) from the historical positioning data set , the value of the parameter λ depends on the requirements of positioning accuracy and the indoor environment. Generally, λ is set between 0.5m and 2m. Assuming that there is a historical positioning track Respectively represent the positioning time, u represents the trajectory The number of nodes, the value of u depends on the size of the indoor space, generally set to 5 to 10) and the current trajectory like then called for close trajectories.

Step 5.2, calculate nodes in similar trajectories corresponds to the node in the current track the distance,

After calculating pair by pair, the distance sequence D=<d ₁ ,d ₂ ,...,d _u > can be obtained.

Step 5.3, calculate the similarity between trajectories

The gray relational degree analysis is based on the degree of similarity between the curve formed by the comparison sequence and the curve formed by the reference sequence to determine the degree of correlation between the comparison sequence and the reference sequence. The larger the value. Therefore, the present invention uses the gray correlation degree of the distance sequence to measure the similarity between the current movement trajectory and the historical movement trajectory. Assuming that the reference sequence is C=<c ₁ ,c ₂ ,...,c _u >, the calculation method of the gray relational degree between the distance sequence D (comparison sequence) and the reference sequence C is as follows:

Step 5.3.1, calculate the absolute value of the difference between the corresponding elements of the distance sequence D and the reference sequence C one by one, namely

|DC|＝|c _v -d _v |,v＝1,2,…,u

Step 5.3.2, determine min(|c _v -d _v |) and max(|c _v -d _v |).

Step 5.3.3, calculate the correlation coefficient of each pair of corresponding elements between the distance sequence D and the reference sequence C

where η is called the resolution coefficient, and η∈[0,1]. The smaller the value of η, the stronger the resolving power, and the greater the difference between the correlation coefficients _Φv , and η = 0.5 is often taken in the actual application process.

Step 5.3.4, calculate the mean value of the correlation coefficient of the corresponding elements of the distance sequence D and the reference sequence C

in (also known as the correlation coefficient) reflects the similarity between the distance sequence D and the reference sequence C, The larger is, the more similar the distance sequence D is to the reference sequence C.

Since the element d _v in the distance sequence D represents the anchor point in the current trajectory Points corresponding to the similar trajectory distance, so the smaller the d _v value is, the closer the two are, and it is optimal when d _v =0, at this time Hence the reference sequence C=<0,0,...0>.

Step 5.4, find the most similar similar trajectory

From numerous similar trajectories, three trajectories most similar to the current trajectory are selected, that is, the three trajectories with the largest correlation coefficient values (respectively set to ), the corresponding correlation coefficient The end points corresponding to similar trajectories are

Step 6, adjust the positioning

From the similarity of indoor objects or moving trajectories, it can be seen that the current position of the point to be located should be located near the area where the end point of the most similar trajectories is located. Using the endpoints of the three most similar proximate trajectories The determined triangular area and the initial coordinate estimation p(xi _′ , y _i ′) of the point to be located can determine the position of the point to be located. Methods as below:

Step 6.1, calculate the position of the inscribed circle center O(x _o ,y _o ) of the triangle

Assume

but

Step 6.2, calculate the intersection point of the line segment pO from the initial coordinates of the point to be located to the center of the circle and the side of the triangle

Assume Then judge whether the line segment pO intersects with the sides AB, AC and BC respectively. The calculation of the intersection point of the line segment pO and the line segment AB is taken as an example for illustration below.

if determinant

Then pO coincides with or is parallel to AB.

If △≠0, calculate

If there is 0≤λ≤1 and 0≤μ≤1, then pO intersects AB, the intersection point (xi _′ +λ(x _O -x _i ′), _i ′+λ(y _O -y _i ′)) is is the actual coordinates of the point to be located; if the three points A, B, and C are collinear, then the coordinates of the midpoint of the longest line segment connected in pairs are the actual coordinates of the point to be located; if point p is within the triangle, then the coordinates of point p That is, the actual coordinates of the point to be located.

The present invention has following beneficial effects:

The indoor positioning method proposed by the present invention introduces the association between the current trajectory and the historical trajectory, and improves the positioning accuracy of indoor objects or personnel whose moving trajectory is relatively stable by establishing an association model.

The present invention will be further described in detail below with reference to the accompanying drawings and embodiments, but the indoor positioning method based on RSSI ranging and trajectory similarity of the present invention is not limited to the embodiments.

Description of drawings

FIG. 1 is a flow chart of an indoor positioning method based on RSSI ranging and moving trajectory similarity according to an embodiment of the present invention;

Fig. 2 is an office layout diagram of an embodiment of the present invention;

Fig. 3 is the original RSSI data set box plot of the embodiment of the present invention;

Fig. 4 is a box plot of the RSSI data set processed by the filtering model according to the embodiment of the present invention.

Detailed ways

Referring to Figures 1 to 4, this embodiment is an indoor positioning method based on RSSI ranging and trajectory similarity. The indoor office is selected as the application scene, and the office of 11.8m×8.9m is selected as the positioning area, and the rectangular coordinates are established. system, where the long direction is the abscissa direction, and the width direction is the ordinate direction. Arrange eight Bluetooth base stations numbered X10001~X10007 in the positioning area, and the corresponding coordinates are (-2.4,-0.6), (-6.8,0), (-11.8,-0.8), (-11.8,-4.8), (-9.7,-8.1), (-2.4,-8.1), (-0.6,-4.4). Now place the device to be positioned in the positioning area, and calculate according to the positioning method of the present invention. The implementation points are as follows:

In step a, the collected RSSI data sets are grouped according to the Bluetooth base station, which can be divided into seven groups in this embodiment, and the data distribution of each group is shown in FIG. 3 .

Step b, each grouped RSSI data set is processed by a filtering model, the Gaussian filtering model eliminates some extreme noises, and selects a relatively stable RSSI value, while the Kalman filtering model makes the relatively stable RSSI data set tend to be smoother, The combination of the two greatly reduces signal noise and improves positioning accuracy. See Figure 4 for the data distribution of each group of RSSI datasets processed by the filtering model, and the final RSSI representative values of each Bluetooth base station are shown in Table 1.

Table 1

bluetooth base station X10001 X10002 X10003 X10004 X10005 X10006 X10007 RSSI value -70 -67 -71 -62 -64 -76 -70

Step c, establish an equation based on the theoretical model of wireless signal transmission and the relationship between the distance formula between the point to be located and each base station, and extract six equations from these seven equations to establish a set of associated equations, which can form seven equations Group. Use the method of maximum likelihood estimation to solve the above seven kinds of related equations respectively. The approximate solutions obtained by using the method of maximum likelihood estimation for each group of equations are shown in Table 2, and the average value of the approximate solutions of the seven related equations is calculated. is (xi _′ , y _i ′) ^T = (6.66, 4.03) ^T .

Table 2

equation set First group Second Group The third group Fourth group fifth group The sixth group seventh group average x _i,s ′ 6.93 6.51 6.64 6.69 6.70 6.51 6.96 6.66 y _i,s ' 3.93 4.29 4.07 4.06 3.71 4.11 4.03 4.03

Step d, find (Because there are many similar trajectories, Table 3 only lists the three with the highest similarity, and the rest will not be listed one by one), and calculate the current trajectory<(1.35,0.83),(2.12,1.57),( 3.00, 2.12), (4.00, 2.76), (4.85, 3.67) > distance sequence to each similar trajectory. Because the application scene space is small, the number of nodes in the trajectory is n=5.

table 3

track node 1 node 2 node 3 node 4 node 5 end track 1 (1.63,0.71) (2.21,1.30) (3.30,2.08) (5.15,2.56) (5.28,3.47) (6.02,4.48) track 2 (1.35,0.13) (2.56,1.72) (3.21,2.55) (4.43,3.07) (5.37,3.85) (5.65,4.94) track 3 (1.63,1.03) (2.15,1.98) (3.44,2.18) (3.58,2.98) (4.96,4.17) (5.06,5.06)

Step e, calculating the similarity between the current trajectory and all similar trajectories, selecting three similar trajectories with the highest similarity, and delimiting the triangular area where the actual position of the positioned device is located by the end of the trajectory. Trajectory 1, Trajectory 2, and Trajectory 3 are the three closest trajectories with the highest similarity, and their distance sequences and similarity values are shown in Table 4. Therefore, the actual position of the positioned device is on the side of the triangle determined by the end points (6.02, 4.48), (5.65, 4.94), and (5.06, 5.06) of the three most similar trajectories.

Table 4

track distance series Similarity track 1 <0.305,0.270,0.302,1.167,0.474> 0.844 track 2 <0.700,0.381,0.479,0.530,0.554> 0.844 track 3 <0.692,0.414,0.444,0.474,0.512> 0.901

Step f, the initial coordinate estimation of the positioned device Combining with the triangular area delineated by its approximate trajectory, the position of the positioned device in the actual coordinate system can be determined. The intersection point (5.92, 4.60) of the line connecting the initial coordinate estimation (6.66, 4.03) and the center of the triangle inscribed circle (5.60, 4.85) and the triangle side (5.92, 4.60) is used as the position of the positioned device in the Cartesian coordinate system.

The above are only preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention Inside.

Claims (8)

1. An indoor positioning method based on RSSI ranging and track similarity, characterized in that, comprising: Step 1, obtain the RSSI data sets of each Bluetooth base station at the location, and group the RSSI data sets according to the signal transmitting base stations corresponding to the collected RSSI data; Step 2, each grouped RSSI data set is screened by a Gaussian filter model, and smoothed by a Kalman filter model, and the smoothed data set is averaged; Step 3, calculate the distance from the point to be located to each base station according to the theoretical model of wireless signal transmission; establish a group of associated equations from the projection distance equation from the point to be located to the positions of each base station; Step 4, use the maximum likelihood estimation method to find the approximate solution of the associated equations; Step 5. Find all the similar trajectories of the current trajectory according to the obtained approximate solution, calculate the distance between the nodes in the current trajectory and the corresponding nodes in the similar trajectories, and calculate the distance sequence between the current trajectory and each similar trajectory; calculate the current trajectory and For the similarity of all similar trajectories, select three similar trajectories with the highest similarity, and delineate the triangular area where the actual position of the positioned device is located by the end points of their trajectories; Step 6, combining the triangular areas determined by the end points of the three most similar trajectories and the initial coordinate estimation of the point to be located to determine the position of the point to be located. 2. the indoor positioning method based on RSSI ranging and track similarity according to claim 1, it is characterized in that, each group of RSSI dataset after grouping is screened by Gaussian filter model, the RSSI data that j base station place group chooses set Wherein, rssi _{j, d, k} represent the kth signal strength indication value collected when the distance from base station j is d; Indicates the mean value of each group of RSSI data sets collected during the positioning data collection period; σ represents the standard deviation of each group of RSSI data sets; n represents the number of rssi values selected in each group; δ represents the significant level, which is small the probability value of Indicates that the probability value is When is the critical value in the normal distribution graph, its value can be found in the normal distribution table. 3. The indoor positioning method based on RSSI ranging and track similarity according to claim 1, wherein said step 3 comprises: Assuming that the initial coordinates of the point to be located are estimated to be (xi _′ , y _i ′), the average value of the data set obtained in step 2 Substituting into the theoretical model of wireless signal transmission, the distance from point p to base station j is d _j ; Assuming that the relative height difference between the terminal to be positioned and the indoor base station is h ₀ , the projection distance equation from the point to be positioned to each base station is obtained by the cosine law, and the following correlation equations are established: The point (x _j , y _j ) represents the coordinates of the base station j. 4. The indoor positioning method based on RSSI ranging and track similarity according to claim 3, wherein said step 4 comprises: A combination method is adopted for the equations in the equation group ①, and j-1 equations are selected from the j equations to form a new equation group, and a total of j types of new equation groups are formed, and the solution of the new sth equation group is The structure is as follows: in Finally, the average value of all j new equation system solutions is calculated as the initial coordinate estimation of the point to be located . 5. The indoor positioning method based on RSSI ranging and trajectory similarity according to claim 4, wherein said step 5 includes: Step 5.1, find similar trajectories Find a point set P={(x,y)||xx _i ′|<λ,| _yy i ′|<λ _} that is closer to the initial position estimate (xi ′, y _i ′) from the historical positioning data set , assuming that there is a historical positioning track and current trajectory like then called for ; among them, the setting of the parameter λ depends on the requirements of positioning accuracy and the indoor environment. Generally, the higher the accuracy requirement or the narrower the indoor space, the smaller the value of λ; t ₁ ,...,t _u+1 represent Positioning time; u represents trajectory The number of nodes, the value of u also depends on the size of the indoor space, generally, the smaller the indoor space, the smaller the value of u; Step 5.2, calculate nodes in similar trajectories corresponds to the node in the current track the distance After calculating pair by pair, the distance sequence D=<d ₁ ,d ₂ ,...,d _u > can be obtained. Step 5.3, calculate the similarity between trajectories Use the gray correlation degree of the distance sequence to measure the similarity between the current movement trajectory and the historical movement trajectory; assuming that the reference sequence is C=<c ₁ ,c ₂ ,…,c _u >, then the gray relation between the distance sequence D and the reference sequence C The calculation method of the correlation degree is as follows: Step 5.3.1, calculating the absolute value of the difference between the corresponding elements of the distance sequence D and the reference sequence C one by one; |DC|＝|c _v -d _v |,v＝1,2,…,u Step 5.3.2, determine min(|c _v -d _v |) and max(|c _v -d _v |); Step 5.3.3, calculate the correlation coefficient of each pair of corresponding elements between the distance sequence D and the reference sequence C where η is called the resolution coefficient, and η∈[0,1]; Step 5.3.4, calculate the mean value of the correlation coefficient of the corresponding elements of the distance sequence D and the reference sequence C in, correlation coefficient The larger is, the more similar the distance sequence D is to the reference sequence C. Step 5.4, find the most similar similar trajectory From many similar trajectories, three trajectories most similar to the current trajectory are selected, that is, the three trajectories with the largest correlation coefficient values are set to Corresponding correlation coefficient The end points corresponding to similar trajectories are 6. The indoor positioning method based on RSSI ranging and trajectory similarity according to claim 5, wherein said step 6 comprises: Step 6.1, calculate the position of the center O(x _O , y _O ) of the inscribed circle of the triangle; make but Step 6.2, calculating the intersection point of the line segment pO from the initial coordinates of the point to be located to the center of the circle and the side of the triangle; Assume Then judge whether there is an intersection point between the line segment pO and the sides AB, AC and BC; if there is an intersection point, then use the intersection point as the actual coordinate of the point to be located; The coordinates of the midpoint are used as the actual coordinates of the point to be located; if point p is within the triangle, the coordinates of point p are used as the actual coordinates of the point to be located. 7. The indoor positioning method based on RSSI ranging and trajectory similarity according to claim 5, wherein the parameter λ is set between 0.5m and 2m. 8. The indoor positioning method based on RSSI ranging and trajectory similarity according to claim 5, wherein the value range of n is set between 5 and 10.