Enhanced Location Method in Complex Scenarios Based on TDOA

Abstract

The present invention discloses an enhanced positioning method in a complex scenario based on TDOA. The steps are as follows: S1 Group frames according to the positioning data received by the base stations to obtain positioning data frames in which a positioning tag sends a positioning message once and reaches multiple base stations; S2 Determine whether the number of base stations reached meets the requirements; S3 Sort the positioning data frames according to the size of the timestamp information and select reference base stations; S4 Combine the reference base stations pairwise and use them as benchmarks in turn to perform positioning calculations on other base stations in the positioning data frames, respectively obtain candidate data sets under each combined benchmark, and then screen candidate data elements in each candidate data set to obtain a hidden state data set; S5 Combine the hidden state data sets of the previous frame and the current positioning frame, calculate the state transition probability of the elements of the current positioning data set, and screen the state transition probability to obtain a positioning result sequence; S6 Obtain the final positioning result. This method improves the positioning stability by combining historical positions.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless positioning, and particularly relates to an enhanced positioning method based on TDOA in complex scenarios. Background Technique

[0002] UWB technology transmits data through extremely narrow pulses instead of traditional carriers, resulting in extremely high data transmission speed. In addition, UWB technology has the advantages of low system complexity, high information security, strong anti-multipath fading ability, etc., and has become a major highlight in the field of wireless positioning.

[0003] The positioning accuracy of UWB based on TDOA is mainly affected by wireless signal propagation and base station topology. For example, problems such as signal reflection, multipath propagation, and non-line-of-sight (NLOS) will all generate different measurement errors, thus affecting the stability and accuracy of the positioning result. In addition, the number of base stations and the deployed topology will also affect the solution accuracy of the TDOA positioning algorithm. A good topology means that the base stations are not concentrated in one area in space and can be evenly distributed in different azimuth areas, so there are certain requirements for both the number of base stations and the arrangement method.

[0004] Currently, the main means to improve the positioning performance of TDOA in complex scenarios include: non-line-of-sight discrimination of the original positioning data to eliminate NLOS data and topological screening of the received base station data to ensure the stability of positioning. These data processes undoubtedly increase the complexity of the algorithm and the real-time performance of backend calculations, and at the same time will cause a reduction in the amount of positioning data, affecting the smoothness of the positioning trajectory.

[0005] Chinese patent document (CN109041207A) discloses a precise positioning system based on BIM technology that can be used for virtual reality and augmented reality, including: a UWB signal sending module; multiple UWB signal receiving base stations; a data processing module that determines the position of the UWB signal sending module relative to N UWB signal receiving base stations according to the TDOA algorithm; a data storage module; multiple cameras evenly arranged in a construction tunnel; a video monitoring module that receives video information captured by each camera; a BIM module that presets an internal electronic map of the construction tunnel, etc. The BIM converts the position information of the UWB signal sending module relative to N UWB signal receiving base stations into position coordinates on the internal electronic map of the construction tunnel; a virtual reality or augmented reality display device. However, this positioning system cannot solve the difficulties of non-line-of-sight discrimination of wireless signals and base station screening.

[0006] The Chinese patent document (CN113030859A) discloses a UWB indoor positioning method based on time division multiple access. The steps include: setting N base stations BSi in the positioning area; the terminal MS transmits a UWB signal once and records the transmission timestamp TMST; after the base station BSi receives the UWB signal, it records the reception timestamp TBSR1(i); the central base station transmits a UWB signal once and records the transmission timestamp TBST; after the positioning base station receives the UWB signal, it records the reception timestamp TBSR2(i); parameters are used to construct a positioning solution matrix to obtain the coordinates of the terminal MS. This UWB indoor positioning method uses N base stations BSi for combined positioning, which can effectively enhance the positioning accuracy of the terminal MS and achieve load balancing of the base stations; by transmitting the UWB signal twice, the terminal MS is positioned through the TDOA positioning method, reducing the deployment cost of the UWB indoor positioning system. However, this positioning method still has insufficient stability in positioning. Summary of the Invention

[0007] In view of the above problems, the present invention provides an enhanced positioning method based on TDOA in complex scenarios, which can improve the positioning performance of UWB in complex scenarios, solve the difficulties of non-line-of-sight identification of wireless signals and base station screening, and at the same time combine historical location information to improve the stability of positioning.

[0008] To solve the above technical problems, the technical solution of the present invention is as follows: The enhanced positioning method based on TDOA in complex scenarios specifically includes the following steps:

[0009] S1: Frame the positioning data received by the base station to obtain a positioning data frame in which a positioning tag sends a positioning message to multiple base stations.

[0010] S2: Determine whether the number of base stations reached meets the requirements. If it meets, go to step S3; if not, return to step S1.

[0011] S3: Sort the positioning data frames in ascending order according to the timestamp information, and select at least 3 base stations as reference base stations according to the timestamp sorting.

[0012] S4: Combine the reference base stations in pairs and use them as benchmarks in turn to perform positioning calculations on other base stations in the positioning data frame, respectively obtain candidate data sets under each combined benchmark, and then screen the candidate data elements in each candidate data set to obtain a hidden state data set.

[0013] S5: Combine the hidden state data set of the previous frame and the hidden state data set of the current positioning frame, calculate the state transition probability of the elements of the current positioning data set, update the transition probability of the hidden state values in the hidden state data set of the current positioning frame, and screen out the state value with the largest state transition probability to obtain a positioning result sequence.

[0014] S6: Perform weighted fusion processing on the positioning result sequence obtained in step S5 to obtain the final positioning result.

[0015] With the above technical solution, first, different base station combination benchmarks are used to solve the candidate positioning data set, covering the positioning solution set with high reliability under single positioning data. Then, the transition probability between two positioning data is calculated by combining with the historical data set, and the most likely state transition path is screened. Finally, weighted fusion processing is performed to obtain the final positioning result, without the need for optimal topology screening and non-line-of-sight processing of the original positioning data, and at the same time, the positioning stability in complex scenarios can be improved.

[0016] Preferably, the positioning data frame in step S1 includes base station coordinate information and the timestamp information of the positioning information arriving at the base station.

[0017] Preferably, at least 4 base stations arriving at in step S2 meet the requirements.

[0018] Preferably, in step S3, the 3 base stations with the earliest timestamp sorting are selected as reference base stations, denoted as reference base station A, reference base station B, and reference base station C respectively; in step S4, the other base stations in the positioning data frame are traversed with the combinations of reference base station A and reference base station B, reference base station A and reference base station C, and reference base station B and reference base station C as benchmarks for positioning calculation respectively, and candidate data sets result_candidate AB 、candidate data set result_candidate AC and candidate data set result_candidate BC are obtained respectively, and then the candidate data elements are screened in each candidate data set to obtain the hidden state data set.

[0019] Preferably, the specific steps of step S4 are:

[0020] S41: Take the coordinates of reference base station A and reference base station B as a combination, that is, take the midpoint of the coordinates of reference base station A and reference base station B as the coordinate origin, rotate the two base stations of reference base station A and reference base station B to the X-axis, and calculate the translation parameter move_para and the rotation parameter theta;

[0021] S42: Convert all base station coordinates to the new reference coordinate system according to the translation parameter move_para and the rotation parameter theta;

[0022] S43: On the basis of reference base station A and reference base station B, traverse the remaining base stations in turn, and use the chan algorithm for three-base-station combination positioning calculation to obtain the candidate data set result_candidateAB ;

[0023] S44: Repeat steps S41 - S43 with reference base station B and reference base station C as a combination and reference base station C and reference base station A as a combination to obtain the candidate data set result_candidate AC and the candidate data set result_candidate BC ;

[0024] S45: Use the K - means clustering algorithm to screen the candidate data set result_candidate AB 、the candidate data set result_candidate AC and the candidate data set result_candidate BC for candidate data elements whose Euclidean distance between candidate data elements meets the set threshold;

[0025] S46: Restore the candidate data values to the original coordinate system according to the translation parameter move_para and the rotation parameter theta to finally obtain the hidden state data set. Use the clustering method to eliminate candidate values with relatively large deviations in the candidate data set and retain candidate values with relatively high clustering degree as the hidden state data set.

[0026] Preferably, the formulas for calculating the translation parameter move_para and the rotation parameter theta in step S41 are:

[0027]

[0028]

[0029] where (x A , y A ) is the coordinate of reference base station A in the original coordinate system, and (x B , y B ) is the coordinate of reference base station B in the original coordinate system.

[0030] Preferably, when converting all base station coordinates to a new reference coordinate system in step S42, the calculation formula when the reference base station is A is:

[0031]

[0032] where is the coordinate of base station A in the new reference coordinate system.

[0033] Preferably, the specific steps for calculating the state transition probability using the Viterbi algorithm in step S5 are:

[0034] Let the hidden state dataset of the previous positioning frame be loc1_candi{j, j = 1, 2,..., M}, the corresponding state transition probability set be para1{j, j = 1, 2,..., M}, the hidden state dataset of the current positioning frame be loc2_candi{i, i = 1, 2,..., N}, and the corresponding state transition probability set be para2{i, i = 1, 2,..., N}. Then the calculation formula for a single element para2(i) in the state transition probability set para2{i, i = 1, 2,..., N} is as follows:

[0035] dist i (j) = ||(loc1_candi(j) - loc2_candi(i))||2;

[0036] transP i (j) = para1(j) × exp(-abs(dist i (j)) 2 );

[0037] para2(i) = max(transP i {j, j = 1, 2,..., M});

[0038] Where loc1_candi(j), loc2_candi(i), and para1(j) are the single element values in loc1_candi{j, j = 1, 2,..., M}, loc2_candi{i, i = 1, 2,..., N}, and para1{j, j = 1, 2,..., M} respectively. dist i (j) is the Euclidean distance between the i-th element in loc2_candi{i, i = 1, 2,..., N} and the j-th element in loc1_candi{j, j = 1, 2,..., M}. transP i (j) is the state transition probability between the i-th element in loc2_candi{i, i = 1, 2,..., N} and the j-th element in loc1_candi{j, j = 1, 2,..., M}.

[0039] Preferably, the calculation formula for the positioning result loc_out in step S6 is:

[0040]

[0041] Among them, loc2_candi{i, i = 1, 2,..., N} is the hidden state dataset of the current positioning frame, and the corresponding state transition probability set is para2{i, i = 1, 2,..., N}, where N is the number of elements in the dataset and i is the index of the dataset element.

[0042] Preferably, the system of the enhanced positioning method in a complex TDOA scenario includes a positioning server, positioning tags, and multiple base stations. Among them, the positioning tags are used to send UWB positioning data; the base stations receive the positioning data sent by the positioning tags, record and process the arrival timestamp information of the positioning data, and then transmit the positioning data to the positioning server; the positioning server is used to receive the positioning data transmitted by the base stations and run the enhanced positioning method in a complex TDOA scenario to obtain the position information of the positioning tags.

[0043] Compared with the prior art, the beneficial effect of the technical solution of the present invention is that: the enhanced positioning method in a complex TDOA scenario uses different base station combination benchmarks to solve the candidate positioning dataset, covering the positioning solution set with high reliability under single positioning data, and then combines the historical dataset to calculate the transfer probability between two positioning data, screens out the most likely state transfer path, and finally performs weighted fusion processing to obtain the final positioning result, realizing the need not to perform optimal topology screening and non-line-of-sight processing of the original positioning data, and at the same time being able to improve the positioning stability in complex scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 is a flowchart of the enhanced positioning method in a complex TDOA scenario of the present invention;

[0045] Figure 2 is a calculation flowchart of the hidden state dataset in the enhanced positioning method in a complex TDOA scenario of the present invention;

[0046] Figure 3 is a schematic diagram of coordinate transformation in the enhanced positioning method in a complex TDOA scenario of the present invention;

[0047] Figure 4 is a schematic diagram of the system structure of the enhanced positioning method in a complex TDOA scenario of the present invention;

[0048] Figure 5 is a comparison diagram of positioning results between the enhanced positioning method in a complex TDOA scenario of the present invention and the classical chan algorithm in the same scenario. DETAILED DESCRIPTION OF THE INVENTION

[0049] In the following description, specific details such as specific system structures and technologies are presented for illustration rather than limitation, so as to thoroughly understand the embodiments of the present invention. Then, those skilled in the art should clearly understand that the present invention can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid unnecessary details from interfering with the description of the present invention.

[0050] Embodiment: As Figure 1 shown, the enhanced positioning method based on TDOA in complex scenarios of the present invention specifically includes the following steps:

[0051] S1: Frame the positioning data received by the base stations to obtain a positioning data frame in which a positioning tag sends a positioning message to multiple base stations; the positioning data frame in step S1 includes base station coordinate information and timestamp information indicating the time when the positioning message arrives at the base station; S2: Determine whether the number of base stations reached meets the requirements. If it meets the requirements, proceed to step S3; if it does not meet the requirements, return to step S1; the number of base stations reached in step S2 meets the requirements when it is at least 4; record the number of base stations received this time as M. If M is less than 4, the positioning is not solved this time, and return to re-frame the data; if M is greater than or equal to 4, obtain a positioning data frame with M rows and 3 columns, where the first column represents the abscissa of the base station, the second column represents the ordinate of the base station, and the third column represents the timestamp information;

[0052] S3: Sort the positioning data frame in ascending order according to the timestamp information, and select at least 3 base stations as reference base stations according to the timestamp sorting; in this embodiment, in step S3, select the 3 base stations with the earliest timestamp sorting as reference base stations, denoted as reference base station A, reference base station B, and reference base station C respectively; that is, sort the positioning data frame in ascending order according to the timestamp information, that is, adjust the rows of the positioning data frame in ascending order from smallest to largest in the third column, and select the 3 base stations with the earliest timestamp sorting as reference base stations, denoted as reference base station A, reference base station B, and reference base station C respectively;

[0053] S4: Combine the reference base stations in pairs and use them as references in turn to traverse other base stations in the positioning data frame for positioning calculation, respectively obtain candidate data sets under each combined reference, and then screen candidate data elements in each candidate data set to obtain a hidden state data set;

[0054] In this embodiment, in step S4, use the combination of reference base station A and reference base station B, the combination of reference base station A and reference base station C, and the combination of reference base station B and reference base station C as references in turn to traverse other base stations in the positioning data frame for positioning calculation, and respectively obtain candidate data sets result_candidate under each combined reference AB, candidate dataset result_candidate AC and candidate dataset result_candidate BC , and then filter the candidate data elements in each candidate dataset to obtain the hidden state dataset; in each candidate dataset, use the clustering method to eliminate the candidate values with relatively large deviations in the candidate dataset, and retain the candidate values with relatively high clustering degree as the hidden state dataset;

[0055] The specific steps of step S4 are as follows:

[0056] S41: Take the coordinates of reference base station A and reference base station B as a combination, that is, take the midpoint of the coordinates of reference base station A and reference base station B as the coordinate origin, rotate the two base stations of reference base station A and reference base station B to the X-axis, and calculate the translation parameter move_para and the rotation parameter theta;

[0057] S42: Convert all base station coordinates to the new reference coordinate system according to the translation parameter move_para and the rotation parameter theta;

[0058] S43: On the basis of reference base station A and reference base station B, sequentially traverse the remaining base stations, and use the classic chan algorithm to perform three-base-station combination positioning calculation to obtain the candidate dataset result_candidate AB ;

[0059] S44: Take reference base station B and reference base station C as a combination and reference base station C and reference base station A as a combination respectively, and repeat steps S41 to S43 to obtain the candidate dataset result_candidate AC and candidate dataset result_candidate BC ;

[0060] S45: Use the K-means clustering algorithm to filter the candidate dataset result_candidate AB , candidate dataset result_candidate AC and candidate dataset result_candidate BC to select the candidate data elements whose Euclidean distance between candidate data elements meets the set threshold;

[0061] S46: Restore the candidate data values to the original coordinate system according to the translation parameter move_para and the rotation parameter theta, and finally obtain the hidden state dataset;

[0062] In this embodiment, as Figure 2 shown, take the AB combination benchmark as an example for illustration;

[0063] (1) Take the midpoint of the coordinates of reference base station A and reference base station B as the coordinate origin, rotate the two base stations of reference base station A and reference base station B onto the X-axis, and calculate the translation parameter move_para and the rotation parameter theta;

[0064] Specifically, the schematic diagram of coordinate system conversion is as Figure 3 shown. The original coordinate system is OXY, and the coordinate system after translation and rotation is O′X′Y′. The specific calculation methods of the translation parameter move_para and the rotation parameter theta are as follows:

[0065]

[0066]

[0067] Among them, (x A , y A ) is the coordinate of reference base station A in the original coordinate system OXY, and (x B , y B ) is the coordinate of reference base station B in the original coordinate system OXY;

[0068] (2) According to the translation parameter move_para and rotation parameter theta calculated in step (1), convert all base station coordinates to the new reference coordinate system O′X′Y′;

[0069] The specific calculation formula is illustrated by taking reference base station A as an example:

[0070]

[0071] Among them, is the coordinate of base station A in the new reference coordinate system O′X′Y′;

[0072] (3) Based on reference base station A and reference base station B, sequentially traverse the remaining base stations one by one, and use the classic chan algorithm to perform three-base-station combined positioning and calculation to obtain the candidate data set result_candidate AB ;

[0073] In this embodiment, taking the number of received base stations M = 6 as an example for illustration, on the premise of taking reference base station A and reference base station B as the reference, each time a remaining reference base station is traversed, then M - 2, that is, 4 groups of base station combinations can be obtained, and the number of elements in the obtained candidate data set result_candidate AB is 4;

[0074] (4) Use the K-means clustering algorithm to screen the candidate data set result_candidateAB Candidate data elements whose Euclidean distance between them satisfies a set threshold; according to different positioning targets, the threshold of the Euclidean distance can be dynamically adjusted. In this embodiment, taking personnel positioning as an example, the default set threshold is 2 meters;

[0075] (5) Restore the candidate data values to the original coordinate system OXY according to the translation parameter move_para and the rotation parameter theta to obtain the final hidden state dataset loc_candi AB ;

[0076] Since the coordinates of the base stations were converted before the positioning calculation, all the solutions in the candidate dataset result_candidate AB are in the new coordinate system O′X′Y′ and need to be restored to the original coordinate system OXY;

[0077] Perform the same processing on the combination of reference base station A and reference base station C and the combination of reference base station B and reference base station C according to the processing method with the combination of reference base station A and reference base station B as the reference to obtain the hidden state dataset loc_candi AC and the hidden state dataset loc_candi BC , and then merge these three datasets into one hidden state dataset loc_candi;

[0078] S5: Combine the hidden state dataset of the previous frame and the hidden state dataset of the current positioning frame, use the Viterbi algorithm to calculate the state transition probability, combine the transition probability of the hidden state values in the hidden state dataset of the previous frame, update the transition probability of the hidden state values in the hidden state dataset of the current positioning frame, and screen out the state value with the largest transition probability to obtain the positioning result sequence; the specific calculation method of the state transition probability in step S5 is as follows:

[0079] Let the hidden state dataset of the previous positioning frame be loc1_candi{j, j = 1, 2,..., M}, the corresponding state transition probability set be para1{j, j = 1, 2,..., M}, the hidden state dataset of the current positioning frame be loc2_candi{i, i = 1, 2,..., N}, and the corresponding state transition probability set be para2{i, i = 1, 2,..., N}. Then the calculation formula for a single element para2(i) in the state transition probability set para2{i, i = 1, 2,..., N} is as follows:

[0080] dist i (j) = ||(loc1_candi(j) - loc2_candi(i))||2;

[0081] transP i(j) = para1(j) × exp(-abs(dist i (j)) 2 );

[0082] para2(i) = max(transP i {j, j = 1, 2,..., M});

[0083] where loc1_candi(j), loc2_candi(i), and para1(j) are the single - element values in loc1_candi{j, j = 1, 2,..., M}, loc2_candi{i, i = 1, 2,..., N}, and para1{j, j = 1, 2,..., M} respectively, dist i (j) is the Euclidean distance between the i - th element in loc2_candi{i, i = 1, 2,..., N} and the j - th element in loc1_candi{j, j = 1, 2,..., M}, and transP i (j) is the state transition probability between the i - th element in loc2_candi{i, i = 1, 2,..., N} and the j - th element in loc1_candi{j, j = 1, 2,..., M};

[0084] S6: Perform weighted fusion processing on the positioning result sequence obtained in step S5 to obtain the final positioning result; the specific calculation formula for the positioning result loc_out is:

[0085]

[0086] where loc2_candi{i, i = 1, 2,..., N} is the hidden state data set of the current positioning frame, the corresponding state transition probability set is para2{i, i = 1, 2,..., N}, N is the number of elements in the data set, and i is the index of the data set element.

[0087] As Figure 4 shown, the schematic diagram of the UWB TDOA positioning system provided by the embodiment of the present invention. The system includes at least one positioning tag, multiple base stations, and a positioning server. Among them, the positioning tag is used to send UWB positioning data; the base station receives the positioning data sent by the positioning tag, records and processes the arrival timestamp information of the positioning data, and then transmits the positioning data to the positioning server; the positioning server is used to receive the positioning data transmitted by the base station and run the enhanced positioning method in complex TDOA scenarios to obtain the position information of the positioning tag.

[0088] To verify the enhanced positioning method in complex scenarios based on TDOA described in this embodiment, a comparative experiment was conducted based on the data collected in a specific project. As Figure 5 shown, 8 base stations were deployed at the engineering project site, and personnel carried positioning tags and walked continuously within the positioning area to collect raw positioning data. For the data collected this time, the enhanced positioning method described in this embodiment and the classic chan algorithm with base station screening were respectively used for positioning calculation to obtain the corresponding positioning trajectories for each method. Through Figure 5 it can be seen that in such a complex engineering site, the chan algorithm will have jumps of several meters or even more than ten meters in some differences, with insufficient stability, while the enhanced positioning method described in this embodiment has better performance and smoother trajectories.

[0089] In the above embodiments, the descriptions of the respective embodiments have their own emphases. For parts not detailed or recorded in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0090] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in the present invention can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

Claims

1. An enhanced positioning method in complex scenarios based on TDOA, characterized in that The specific steps are as follows: S1: Frame the positioning data received by the base station to obtain a positioning data frame in which a positioning tag sends a positioning message to multiple base stations; S2: Determine whether the number of base stations reached meets the requirements. If it meets, go to step S3; if not, return to step S1; S3: Sort the positioning data frames in ascending order according to the timestamp information, and select at least 3 base stations as reference base stations according to the timestamp sorting; S4: Combine the reference base stations in pairs and use them as benchmarks in turn to perform positioning calculations on other base stations in the positioning data frame, respectively obtain candidate data sets under each combined benchmark, and then screen the candidate data elements in each candidate data set to obtain a hidden state data set; S5: Combine the hidden state data set of the previous frame and the hidden state data set of the current positioning frame, calculate the state transition probability of the elements of the current positioning data set, update the transition probability of the hidden state values in the hidden state data set of the current positioning frame, and screen out the state value with the largest state transition probability to obtain a positioning result sequence; S6: Perform weighted fusion processing on the positioning result sequence obtained in step S5 to obtain the final positioning result.

2. The enhanced positioning method based on TDOA in complex scenarios according to claim 1, wherein The positioning data frame in step S1 includes the base station coordinate information and the timestamp information when the positioning information reaches the base station; the number of base stations reached in step S2 meets the requirements if it is at least 4.

3. The enhanced positioning method in a complex scenario based on TDOA according to claim 1, characterized in that, In step S3, the 3 base stations at the front of the timestamp sorting are selected as reference base stations, which are respectively denoted as reference base station A, reference base station B, and reference base station C; In the step S4, taking the reference base station A and the reference base station B as a combination, the reference base station A and the reference base station C as a combination, and the reference base station B and the reference base station C as a combination in sequence as a reference to traverse other base stations in the positioning data frame for positioning calculation, and respectively obtaining candidate data sets result_candidate under each combination reference AB , candidate data set result_candidate AC and candidate data set result_candidate BC , and then screening candidate data elements in each candidate data set to obtain a hidden state data set.

4. The enhanced positioning method based on TDOA in complex scenarios according to claim 3, characterized in that, The specific steps of step S4 are as follows: S41: Take the coordinates of reference base station A and reference base station B as a combination, that is, take the midpoint of the coordinates of reference base station A and reference base station B as the coordinate origin, rotate the two base stations of reference base station A and reference base station B to the X-axis, and calculate the translation parameter move_para and the rotation parameter theta; S42: Convert the coordinates of all base stations to the new reference coordinate system according to the translation parameter move_para and the rotation parameter theta; S43: Based on the reference base stations A and B, sequentially traverse the remaining base stations, and use the chan algorithm to perform three-base-station combined positioning calculation to obtain the candidate dataset result_candidate AB ; S44: Repeat steps S41 - S43 with reference base station B and reference base station C as a combination and reference base station C and reference base station A as a combination to obtain candidate dataset result_candidate AC and candidate dataset result_candidate BC ; S45: Use the K-means clustering algorithm to screen the candidate data set result_candidate AB and the candidate data set result_candidate AC and the candidate data elements in the candidate data set result_candidate BC whose Euclidean distance between candidate data elements meets the set threshold; S46: Restore the candidate data values to the original coordinate system according to the translation parameter move_para and the rotation parameter theta, and finally obtain a hidden state data set.

5. The enhanced positioning method based on TDOA in complex scenarios according to claim 4, wherein , The formulas for calculating the translation parameter move_para and the rotation parameter theta in step S41 are: Among them, (x A , y A ) is the coordinate of reference base station A in the original coordinate system, and (x B , y B ) is the coordinate of reference base station B in the original coordinate system.

6. The enhanced positioning method based on TDOA in complex scenarios according to claim 5, characterized in that, When converting the coordinates of all base stations to the new reference coordinate system in step S42, the calculation formula when the reference base station is A is: Among them, is the coordinate of base station A in the new reference coordinate system.

7. The enhanced positioning method based on TDOA in complex scenarios according to claim 6, characterized in that The specific steps of calculating the state transition probability using the Viterbi algorithm in step S5 are: Let the hidden state dataset of the previous positioning frame be loc1_candi{j, j = 1, 2, ..., M}, and the corresponding state transition probability set be par1a{, j = j1, 2, .., M}. The hidden state dataset of the current positioning frame is loc2_candi{i, i = 1, 2, ..., N}, and the corresponding state transition probability set is para2{i, i = 1, 2, ..., N}. Then the calculation formula for a single element para2(i) in the state transition probability set para2{i, i = 1, 2, ..., N} is as follows: dist i (j) = (loc1_candi(j) - loc2_candi(i))2; transP i (j) = para1(j) × exp(-abs(dist i (j)) 2 ); para2(i) = max(transP i {j, j = 1, 2,..., M}); Among them, loc1_candi(j), loc2_candi(i), and para1(j) are the single element values in loc1_candi{j, j = 1, 2,..., M}, loc2_candi{i, i = 1, 2,..., N}, and para1{j, j = 1, 2,..., M}, respectively, and dis i t(j is the Euclidean distance between the i-th element in loc2_candi{i, i = 1, 2,..., N} and the j-th element in loc1_candi{j, j = 1, 2,..., M}, and transP i (j) is the state transition probability between the i-th element in loc2_candi{i, i = 1, 2,..., N} and the j-th element in loc1_candi{j, j = 1, 2,..., M}.

8. The enhanced positioning method based on TDOA in complex scenarios according to claim 4, characterized in that, The calculation formula for the positioning result locout in step S6 is: Where loc2_candi{i, i = 1, 2, ..., N} is the hidden state dataset of the current positioning frame, the corresponding state transition probability set is para2{i, i = 1, 2, ..., N}, N is the number of elements in the dataset, and i is the index of the dataset element.

9. The enhanced positioning method based on TDOA in complex scenarios according to any one of claims 1-8, characterized in that, The system of this enhanced positioning method in a complex TDOA scenario includes a positioning server, positioning tags, and multiple base stations. Among them, the positioning tags are used to send UWB positioning data; the base stations receive the positioning data sent by the positioning tags, record and process the arrival timestamp information of the positioning data, and then transmit the positioning data to the positioning server; the positioning server is used to receive the positioning data transmitted by the base stations and run this enhanced positioning method in a complex TDOA scenario to obtain the position information of the positioning tags.

Images