Direct positioning method, system and device based on aoa and tdoa multi-modal geometric constraints

Abstract

The application discloses a direct positioning method and system based on AOA and TDOA multi-modal geometric constraints, and belongs to the technical field of radio monitoring and passive positioning. The method solves the problem that a traditional TDOA-DPD positioning method needs to divide a high-density calculation grid in a full detection area, realizes double promotion of calculation speed and positioning accuracy by introducing a multi-modal geometric observation boundary, shears a TDOA time difference error band by using an AOA angle measurement error cone, effectively reduces a search area, has the advantages of simple realization and simple deployment, and is suitable for practical production.

Description

Technical Field

[0001] This invention belongs to the field of radio monitoring and passive positioning technology, and relates to a passive positioning method based on spectrum devices, specifically a direct positioning method, system and device based on multimodal geometric constraints of angle of arrival (AOA) and time difference of arrival (TDOA). Background Technology

[0002] With the increasing complexity of the electromagnetic environment, rapid and high-precision localization of non-cooperative radiation sources (such as unauthorized drones, radio stations, and jamming sources) has become a core requirement for radio detection and electronic countermeasures systems. Passive localization technology for non-cooperative targets is a popular research area in both academia and industry, and has been widely applied in practice. Passive localization offers advantages such as strong concealment and low cost, and is widely used in military and civilian fields including electronic reconnaissance, detection of unauthorized drone flights, location of unauthorized radio broadcasts, and search and rescue. Especially in the detection of unauthorized drone flights, traditional radar is limited by the small RCS (Radar Cross Section) and low flight altitude of unauthorized drones, easily leading to certain blind spots. Passive localization technology, due to its versatility, can intercept and locate any signal broadcast by the drone, compensating for the shortcomings of active radar localization.

[0003] Current passive positioning technologies can be divided into two categories based on their implementation characteristics. The first is the traditional two-step positioning method, which is currently the most widely used in engineering applications. This method first intercepts the target signal, then solves the signal to obtain positioning parameters, and finally establishes a system of equations based on the characteristics between the positioning parameters and the target position to solve for the position. Typical two-step positioning methods include those based on Angle of Arrival (AOA), Time Difference of Arrival (TDOA), and Frequency Difference of Arrival (FDOA). Taking the TDOA-based positioning method as an example, multiple stations simultaneously receive the target signal. After determining the master station, the time difference between the target's arrival at each station and its arrival at the master station is calculated. Finally, the time difference is substituted into the solution equations to solve for the target position. The advantages of the two-step positioning method are low computational complexity and good real-time performance. The disadvantage is that the two-step process introduces factors that can lead to positioning errors. The second category is the newer one-step positioning method, also known as the Direct Position Determination (DPD) method. This method eliminates the traditional process of solving for positioning parameters. The process involves acquiring UAV signals, constructing a cost function associated with the signal source location, and then using methods such as exhaustive search to accurately estimate the target location. The DPD method does not require pre-estimation of parameters, but when constructing the cost function, information containing the target signal source location needs to be implicitly embedded, such as AOA, TDOA, FDOA, and RSSI (Received Signal Strength Indication).

[0004] In real-world positioning scenarios, considering the target's multiple motion characteristics, including stationary and maneuvering, a direct positioning method based on cross-correlation accumulation, namely the TDOA-DPD method, is often used. This method requires only one step and can achieve higher accuracy by dividing the grid into smaller parts. However, it typically employs a full-grid blind search strategy, requiring a comprehensive scan of the entire monitoring grid regardless of the quality of the observation data. This results in an extremely high computational load for the TDOA-DPD method, making it difficult to apply in actual production due to hardware performance limitations and real-time requirements. Furthermore, while the TDOA-DPD method improves positioning accuracy compared to the traditional TDOA method and reduces the blind zone to some extent, it does not completely solve the problem of blind zones, especially when the site's geometric layout is irregular, making the unpredictability of the blind zone significant. Traditional two-step methods, such as TDOA or AOA, require extracting intermediate parameters like TDOA or AOA observations from noisy signals. Especially in complex electromagnetic environments and with weak signal strength, large parameter estimation errors can drastically reduce positioning accuracy. For AOA, the positioning error caused by angle measurement increases linearly with distance, leading to divergent and ambiguous positioning results. For TDOA, the long-tail effect of hyperbolic curves can produce ambiguous solutions, also known as blind zones, when the base station's geometry is poor. Traditional TDOA+AOA fusion methods can address the blind zone problem to some extent, but the errors introduced by the two-step method remain unresolved. Therefore, a new positioning method is urgently needed in passive positioning scenarios to effectively address the high computational complexity of traditional methods while maintaining positioning performance and avoiding TDOA blind zones. Summary of the Invention

[0005] This invention aims to provide a direct positioning method based on multimodal geometric constraints of AOA and TDOA, which can solve the problem of high computational complexity of traditional methods, while taking into account positioning performance and avoiding TDOA blind spots. It is particularly suitable for passive positioning scenarios with limited computing resources and high requirements for real-time performance and accuracy.

[0006] To achieve the above objectives, the technical solution adopted by this invention is: a direct positioning method based on AOA and TDOA multimodal geometric constraints, comprising:

[0007] Multiple sites simultaneously acquire target signals emitted by the radiation source; the multiple sites include multiple individual sites, which are used to form a monitoring network deployed in the monitoring area;

[0008] Obtain the AOA observation value of the single site, which is used to characterize the direction of arrival of the target signal;

[0009] Obtain TDOA observations for base station pairs; each base station pair includes two single sites, one of which is a reference site and the other is the remaining single site excluding the reference site;

[0010] An effective grid area is obtained based on the spatial intersection of the AOA angle measurement error cone and the TDOA time difference error band. The AOA angle measurement error cone is a fan-shaped area radiating outwards from the single station as the base point, the estimated distance from the radiation source to the single station as the radius, the direction of arrival as the center line, and the angle error threshold as half an angle. The TDOA time difference error band is the area enclosed by the extended curve. The extended curve is obtained through a branch curve and a distance error threshold. The branch curve is the target branch of the positioning hyperbola falling within the monitoring area. The positioning hyperbola is a hyperbola generated based on the distance difference with the base station pair as the focus. The distance difference is the distance difference from the radiation source to the base station pair, obtained by converting the TDOA observation values ​​of the base station pair.

[0011] Based on the geometric positions of candidate target points within the effective grid area, the theoretical time difference of the target signal arriving at the base station pair is calculated, and the cross-correlation function value of the target signal received by the base station pair is calculated; effective cross-correlation function values ​​are selected based on the coherence coefficient threshold, and the effective cross-correlation function values ​​are accumulated to obtain the coherence accumulation result;

[0012] Within the coordinate space corresponding to the effective grid region, the global maximum value of the coherent cumulative cost function is searched, and the coordinate point corresponding to the global maximum value is substituted into the direct localization algorithm to output the final estimated coordinates of the radiation source target; wherein, the coherent cumulative cost function is the global peak function of the coherent cumulative result corresponding to the coordinate point, and the coordinate point is a point within the coordinate space corresponding to the effective grid region.

[0013] The direct positioning method based on AOA and TDOA multimodal geometric constraints provided by this invention is suitable for passive positioning scenarios. It retains the high accuracy and anti-interference capability of the TDOA-DPD method under low signal-to-noise ratio conditions, while overcoming the huge computational burden caused by full-grid search. Simultaneously, it eliminates positioning blind spots based on the complementarity of multimodal data. By dividing the pre-positioning region through multimodal geometric constraints to reduce the TDOA-DPD search grid area, the computational burden of the algorithm is greatly reduced, making it particularly valuable for applications requiring large-scale positioning. Furthermore, the pre-positioning region utilizes the device's angle measurement information of the target, avoiding the blind spot effect caused by the traditional TDOA baseline. Based on the angle measurement error angle and TDOA observations, the pre-positioning region is simple to implement and closely resembles real-world scenarios. It exhibits strong robustness to varying electromagnetic environments and is unaffected by the target's maneuvering or stationary state.

[0014] The proposed method achieves an order-of-magnitude improvement in computational efficiency by introducing physical error boundaries for observations and constructing dual-modal geometric constraints for AOA and TDOA. The traditional exhaustive full-grid search is transformed into a sparse search over high-confidence regions. Without sacrificing the original accuracy of the TDOA-DPD positioning method, most low-confidence regions are eliminated, significantly reducing computational power consumption. By fusing the directionality of AOA observations and the time difference characteristics of TDOA observations, a high-confidence pre-positioning region is defined using an angle measurement error cone to geometrically divide the time difference error band, eliminating positioning blind spots, enhancing the robustness of the positioning system, and avoiding the fuzzy solution problem inherent in traditional TDOA methods.

[0015] According to embodiments of the present invention, the present invention can be further optimized, and the optimized technical solution is as follows:

[0016] In one preferred embodiment, the acquisition of the AOA observations specifically includes:

[0017] The target signal is received by the array antenna of the single site, the target signal is processed by angle measurement, the frequency domain is transformed by discrete Fourier transform, and the AOA observation value representing the direction of arrival of the target signal is output by subspace analysis.

[0018] The acquisition of the TDOA observations specifically includes:

[0019] The base station synchronously receives the target signal emitted by the radiation source, converts it to the frequency domain through discrete Fourier transform, and obtains the frequency domain signal; based on the phase characteristics of the cross power spectrum of the frequency domain signal, the base station calculates the time delay difference between receiving the target signal and the target signal as the TDOA observation value.

[0020] The solution provided by this invention can effectively suppress noise and multipath interference, improve the accuracy of phase difference measurement and time difference measurement, adapt to high-resolution direction finding algorithms, reduce system synchronization requirements, enhance broadband signal processing capabilities and system robustness, and efficiently integrate with cross-correlation and coherent accumulation processes, thereby improving the accuracy and stability of AOA and TDOA observations.

[0021] In one preferred embodiment, obtaining the estimated distance from the radiation source to the single site specifically includes:

[0022] Based on the received strength of the target signal, the estimated distance from the radiation source to the single site is obtained according to the spatial propagation loss model.

[0023] In one preferred embodiment, the generation of the effective grid region specifically includes:

[0024] Construct a binarized sparse constraint matrix, which is the projection of a binarized mask function onto a discrete space; the discrete space is a set of candidate target points; the projection is the operation of mapping the output value of the binarized mask function at the candidate target points to the corresponding element of the binarized sparse constraint matrix.

[0025] The binary mask function The range of values ​​is {0, 1}, and the mapping rule satisfies:

[0026] When the candidate target point falls into the spatial intersection of the AOA angle measurement error cone and the TDOA time difference error band... ;

[0027] When the candidate target point does not fall within the spatial intersection of the AOA angle measurement error cone and the TDOA time difference error band ;

[0028] in, The coordinates of the candidate target point are given.

[0029] The solution provided by this invention utilizes the AOA (Angle Angle Error Cone) to cut the TDOA (Time Difference Error Band). The intersection of the angle error cone and the time difference error band constitutes the constraint boundary. Further binarization of this constraint boundary yields a sparse geometric constraint matrix. The construction of this binarized sparse constraint matrix simplifies computation and reduces computational complexity.

[0030] In one preferred embodiment, the relationship between the candidate target point falling into the AOA angle measurement error cone is as follows:

[0031] ;

[0032] Where mod 360 represents taking the modulus of the angle, with a value within 360°; For the first Location coordinates of a single site; For the first Angle of arrival obtained from actual measurements at a single site; The angle error threshold is... It is a positive integer. ; The number of the multiple sites;

[0033] The specific formula for determining whether the candidate target point falls within the TDOA time difference error band is as follows:

[0034] ;

[0035] in, For the first Location coordinates of a single site; For the first Location coordinates of a single site; , , ; For the candidate target point obtained by actual measurement to reach the first Single station and arrival The time difference between individual sites, i.e., the time difference between the base stations TDOA observations; The speed of light; For the radiation source to the base station pair The distance difference, the first Each single site serves as a reference site; The distance error threshold is defined as follows.

[0036] Compared to the traditional TDOA-DPD method, which typically uses AOA angle measurement information directly to construct a pre-positioning region, the direct positioning method proposed in this invention utilizes AOA and TDOA observations to construct angle measurement error cones and time difference error bands, respectively. This transforms the abstract statistical errors in the traditional TDOA-DPD method into concrete geometric boundaries, i.e., positioning region search is performed based on physical error boundaries, constructing a fine and accurate computational grid region. The method fully leverages the hyperbolic characteristics of TDOA observations to construct the time difference error band and uses AOA observations to trim it, thus solving the problem of the inherent positioning blind zone in the traditional TDOA-DPD positioning method without affecting the positioning accuracy of TDOA-DPD.

[0037] In one preferred embodiment, the angle error threshold The distance error threshold ;in, The standard deviation of the azimuth angle. This represents the standard deviation of the time difference.

[0038] The present invention employs the Raida criterion to remove outliers from the observations, effectively suppressing the impact of abnormal noise on subsequent coherent accumulation and positioning calculations, thereby improving the reliability of the observations and the robustness of the system.

[0039] Based on the same concept, the present invention also provides a direct positioning system based on AOA and TDOA multimodal geometric constraints, comprising:

[0040] A broadband signal acquisition module is used to synchronously acquire target signals emitted by radiation sources from multiple sites. The multiple sites include multiple single sites, which are used to form a monitoring network deployed in the monitoring area.

[0041] A multimodal parameter estimation module is used to obtain the AOA observation value of the single site, which is used to characterize the direction of arrival of the target signal; and to obtain the TDOA observation value of the base station pair; each base station pair includes two single sites, one of which is a reference site and the other is the remaining single site excluding the reference site.

[0042] The effective grid area acquisition module is used to obtain an effective grid area based on the spatial intersection of the AOA angle measurement error cone and the TDOA time difference error band. The AOA angle measurement error cone is a fan-shaped area radiating outwards from the single station as the base point, the estimated distance from the radiation source to the single station as the radius, the direction of arrival as the center line, and the angle error threshold as half an angle. The TDOA time difference error band is the area enclosed by the extended curve. The extended curve is obtained through a branch curve and a distance error threshold. The branch curve is the target branch of the positioning hyperbola falling within the monitoring area. The positioning hyperbola is a hyperbola generated based on the distance difference with the base station pair as the focus. The distance difference is the distance difference from the radiation source to the base station pair, obtained by converting the TDOA observation values ​​of the base station pair.

[0043] The coherent accumulation module is used to calculate the theoretical time difference of the target signal arriving at the base station pair based on the geometric position of the candidate target points in the effective grid area, and to calculate the cross-correlation function value of the target signal received by the base station pair; to filter out the effective cross-correlation function values ​​based on the coherence coefficient threshold, and to accumulate the effective cross-correlation function values ​​to obtain the coherent accumulation result;

[0044] The target location determination module is used to search for the global maximum value of the coherent cumulative cost function in the coordinate space corresponding to the effective grid area, substitute the coordinate point corresponding to the global maximum value into the direct positioning algorithm, and output the final estimated coordinates of the radiation source target; wherein, the coherent cumulative cost function is the global peak function of the coherent cumulative result corresponding to the coordinate point, and the coordinate point is a point in the coordinate space corresponding to the effective grid area.

[0045] This invention provides a direct localization system based on multimodal geometric constraints of AOA and TDOA, targeting non-cooperative targets, including targets with varying motion states such as unauthorized UAVs and fixed black base stations, which exhibit low speed, slow movement, and small size. The solution provided by this invention leverages the characteristic that TDOA is unaffected by the target's maneuvering or stationary state, selecting TDOA measurements as the signal source for modeling. It utilizes the AOA angle measurement error cone to cut off the TDOA time difference error band, effectively reducing the search area. This system is simple to implement and deploy, making it suitable for practical production.

[0046] Based on the same concept, the present invention also provides an electronic device, including a memory, a processor, and a computer program or instructions stored in the memory, wherein the processor executes the computer program or instructions to implement the direct positioning method based on AOA and TDOA multimodal geometric constraints as described above.

[0047] Compared with the prior art, the beneficial effects of the present invention are:

[0048] The direct positioning method proposed in this invention, based on multimodal geometric constraints of AOA and TDOA, solves the problem that the traditional TDOA-DPD positioning method requires dividing a high-density computational grid in the entire detection area. By introducing multimodal geometric observation boundaries, it achieves a dual improvement in computational speed and positioning accuracy. It uses the AOA angle measurement error cone to cut the TDOA time difference error band, effectively reducing the search area. It has the advantages of simple implementation and simple deployment, and is suitable for actual production. Attached Figure Description

[0049] Figure 1 This is a flowchart of a direct localization method based on multimodal geometric constraints of AOA and TDOA according to an embodiment of the present invention;

[0050] Figure 2 This is a region partitioning diagram based on a multimodal geometric constraint region according to an embodiment of the present invention;

[0051] Figure 3 This is a region partitioning diagram based on multimodal geometric constraints for irregular base station deployment according to an embodiment of the present invention. Detailed Implementation

[0052] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other.

[0053] TDOA (Time Difference of Arrival) Positioning Principle: TDOA positioning, also known as hyperbolic positioning, uses distance difference as the product of time difference and electromagnetic wave velocity in non-cooperative target positioning. The target's possible location lies on a hyperbola with two receiving stations as foci. If there are three receiving stations, two hyperbolic trajectories can be formed, and the target's location is the intersection of these two hyperbolas. The two key steps in traditional TDOA positioning methods are: calculating the time difference and solving the hyperbolic equation.

[0054] Direct Positioning Device (DPD) technology principle: Unlike traditional two-step positioning methods that involve position parameter estimation and equation solving, the DPD method combines observations from all stations and directly solves for the target's position in a single step. The positioning accuracy of this one-step method has been proven superior to traditional two-step methods in multiple studies because it avoids positioning errors introduced during multi-step processes. After acquiring non-cooperative target signals, the DPD method constructs a cost function associated with the signal source's location and then uses methods such as exhaustive search to achieve accurate estimation of the UAV's position. Modeling the cost function requires considering the target signal modeling; the information carried by the wireless signal includes amplitude, time delay, and Doppler shift. Due to significant electromagnetic interference in real-world scenarios, amplitude measurement is greatly affected, and the low speed of the UAV leads to low Doppler shift accuracy. Therefore, the TDOA-based DPD method is widely used in practical scenarios.

[0055] The principle of the TDOA-DPD method: assuming the carrier frequency of the target signal is... The distance from the target to the station is , establish the first The formula for the received signal model of a single station is as follows:

[0056] (1);

[0057] in, For the first Sites The received signal at any given moment; Indicates the amplitude of the signal; This indicates the transmission process of the transmitted signal after a time delay. Indicates time delay. ; Represents the speed of light; Indicates phase shift, ; Indicates frequency offset; , For the number of sites; This represents Gaussian white noise during signal transmission.

[0058] As can be seen from formula (1), the effect of time delay is introduced in signal modeling, that is, implicit calculation of TDOA is included.

[0059] In this embodiment of the invention, the location of the target signal is determined by solving for the unknown quantity using the maximum likelihood function, where the error between the established received signal model and the actual received signal is minimized. The positioning result is then obtained. It is expressed as follows:

[0060] (2);

[0061] in, This indicates the duration of the received signal; for higher positioning accuracy. The frequency offset caused by the relative motion between the target and the receiving station during maneuvering is considered, and the calculation formula is as follows:

[0062] (3);

[0063] The fast TDOA-DPD method based on AOA-TDOA multimodal geometric constraints provided in this invention is based on the following principle:

[0064] Geometric Construction of Multimodal Error Boundaries: In the initial stage of signal processing, a single site uses an array antenna, and the direction finding method employs amplitude comparison direction finding or phase interferometry algorithm to obtain the AOA angle measurement results. Multiple sites simultaneously receive signals, and the TDOA time difference results are obtained using a cross-correlation method. Due to noise, the AOA angle measurement results and TDOA time difference results do not exhibit ideal ray or hyperbola characteristics in geographic space, but rather represent an error field with a certain probability distribution.

[0065] Constructing the angle measurement error cone: for any positioning point Compared to the first Sites Theoretical azimuth The definition is as follows:

[0066] (4);

[0067] Location point Falling into Angle measurement error cone at each station The mathematical conditions are expressed as follows:

[0068] (5);

[0069] The physical meaning of formula (5) is: the minimum geometric angle between the theoretical line-of-sight direction and the measurement direction of the positioning point is less than or equal to the error threshold. The equivalent relationship of formula (5) is as follows:

[0070] (6);

[0071] in, For the first Angle of arrival obtained from actual measurements at each station; For example, take the angle error threshold. , The standard deviation of the angle refers to the dispersion of azimuth data obtained from multiple direction measurements; the smaller the value, the higher the direction measurement accuracy of the AOA observations. In this embodiment of the invention, The possible value is 3 degrees. The maximum angle error threshold can also be given based on the environment and experience.

[0072] Construction of the time difference hyperbola error band: location points To the Sites Euclidean distance The calculation formula is as follows:

[0073] (7);

[0074] In this embodiment of the invention, the first station is designated as the reference station, and the second station is designated as the reference station. Each site is an auxiliary site. The theoretical distance difference generated by the positioning point. as follows:

[0075] (8);

[0076] Location point Falling into base station Time difference error band The mathematical conditions are expressed as follows:

[0077] (9);

[0078] The physical meaning of formula (9) is: the absolute value of the difference between the theoretical distance difference from the positioning point to the two stations and the observed distance difference must be within the physical error tolerance.

[0079] in, The first one obtained from actual measurement The TDOA value between the station and the reference station, At the speed of light, This is the distance error threshold. In this invention, Desirable , The standard deviation of time difference (TDOA) refers to the dispersion of the received signal time difference between base station pairs obtained from multiple measurements. A smaller standard deviation indicates higher measurement accuracy of the TDOA observation. Its value is determined by the interference level of the application scenario, hardware performance, and signal characteristics. Angle error thresholds and distance error thresholds can be set according to the electromagnetic environment, or calculation logic can be set in the program to automatically and dynamically adjust the errors based on the electromagnetic environment to form different angle measurement error cones and time difference error bands. In this embodiment of the invention, The possible value is 20 nanoseconds.

[0080] The direct positioning method based on AOA-TDOA multimodal geometric constraints provided in this invention relies on calculating the spatial intersection of the angle measurement error cone and the time difference error band. The entire monitoring area is considered as consisting of countless positioning points. The plane formed by these two elements is treated as a filter by the binary mask function. Specifically, the binary mask geometric constraint function is defined. as follows:

[0081] (10);

[0082] A value of 1 indicates that the location point meets the physical constraints and the DPD location method needs to calculate the location point; a value of 0 indicates that the target cannot possibly be at the location point and the DPD does not need to calculate the location point.

[0083] Expanding formula (10) into a system of explicit inequalities, we get the positioning points that the DPD positioning method needs to calculate. All of the following conditions must be met simultaneously:

[0084] 1. AOA constraint (all base stations):

[0085] Definition of the first The location of the base station is Measured relative to candidate targets The corresponding actual azimuth angle is The definition of AOA constraint is:

[0086] (11);

[0087] in, Used to calculate the angle between candidate targets and the base station, all of which are less than a threshold. The candidate targets form the region of the AOA angle measurement error cone, which geometrically appears as a fan-shaped region with the base station as the vertex, i.e., a set. The region to which it belongs. For example... Figure 2 As shown, the yellow triangles represent base stations.

[0088] 2. TDOA Constraints (All Base Station Pairs)

[0089] In this embodiment of the invention, base station 1 is used as the reference site, and other base stations... Together with base station 1, they form a base station pair. This indicates that the candidate target point obtained from actual measurement has reached the base station. The time difference with arrival at base station 1. The TDOA constraint is defined as:

[0090] (12);

[0091] The first two square roots represent points respectively. to base station Geometric distance from base station 1 This represents the physical distance difference converted from the time difference. It is the speed of light. Geometrically, it is represented as a hyperbolic strip-shaped region with the two base stations as foci, i.e., a set. Region. The reference site selection can be adjusted according to the actual scenario.

[0092] like Figure 1 The diagram shows the overall flowchart of the direct localization method based on multimodal geometric constraints of AOA and TDOA provided in this embodiment of the invention. The binarized sparse constraint matrix is ​​the projection of the binarized mask function M(x,y) onto the discrete space. Its sparsity is determined by the spatial proportion of the geometric constraint conditions (the intersection region of AOA and TDOA), aiming to provide computational guidance for subsequent coherent accumulation.

[0093] like Figure 3 As shown, in real-world deployments, especially at the city level, it is often difficult to achieve a perfectly regular geometric deployment. Therefore, it is necessary to consider the effectiveness of the method provided in this embodiment under irregular deployment conditions. Figure 3 The example also uses four stations, but the geometric deployment of the stations is irregular. Using the traditional TDOA-DPD positioning method to search the entire area requires calculating all grid points within x=-2500:2500 and y=-2500:2500. However, the method provided in this embodiment of the invention can quickly segment the effective positioning area. Figure 3 The effective DPD area (marked in pink) significantly reduces the computational load and positioning blind spots of the traditional TDOA-DPD method.

[0094] This invention also provides a direct positioning system based on AOA and TDOA multimodal geometric constraints, including:

[0095] A broadband signal acquisition module is used to synchronously acquire target signals emitted by radiation sources from multiple sites. The multiple sites include multiple single sites, which are used to form a monitoring network deployed in the monitoring area.

[0096] A multimodal parameter estimation module is used to obtain the AOA observation value of the single site, which is used to characterize the direction of arrival of the target signal; and to obtain the TDOA observation value of the base station pair; each base station pair includes two single sites, one of which is a reference site and the other is the remaining single site excluding the reference site.

[0097] The effective grid area acquisition module is used to obtain the effective grid area based on the spatial intersection of the AOA angle measurement error cone and the TDOA time difference error band. The AOA angle measurement error cone is a fan-shaped area radiating outwards from the single station as the base point, the estimated distance from the radiation source to the single station as the radius, the direction of arrival as the center line, and the angle error threshold as half an angle. The TDOA error band is the area enclosed by the extended curve. The extended curve is obtained through a branch curve and a distance error threshold. The branch curve is the target branch of the positioning hyperbola falling within the monitoring area. The positioning hyperbola is a hyperbola generated based on the distance difference with the base station pair as the focus. The distance difference is the distance difference from the radiation source to the base station pair, obtained by converting the TDOA observation values ​​of the base station pair.

[0098] The coherent accumulation module is used to calculate the theoretical time difference of the target signal arriving at the base station pair based on the geometric position of the candidate target points in the effective grid area, and to calculate the cross-correlation function value of the target signal received by the base station pair; to filter out the effective cross-correlation function values ​​based on the coherence coefficient threshold, and to accumulate the effective cross-correlation function values ​​to obtain the coherent accumulation result;

[0099] The target location determination module searches for the global maximum value of the coherent cumulative cost function within the coordinate space corresponding to the effective grid area, substitutes the coordinate point corresponding to the global maximum value into the direct positioning algorithm, and outputs the final estimated coordinates of the radiation source target; wherein, the coherent cumulative cost function is the global peak function of the coherent cumulative result corresponding to the coordinate point, and the coordinate point is a point within the coordinate space corresponding to the effective grid area.

[0100] The direct positioning method based on AOA and TDOA multimodal geometric constraints proposed in this invention determines the angle error threshold and time difference error threshold according to the characteristics of the actual scene. For example, the error threshold in complex urban environments is greater than that in suburban environments. An angle error cone is constructed using AOA measurement errors, and a time difference error band is constructed using TDOA measurement errors. The angle error cone is defined as the area formed by extending outwards from a single station as the base point, the direction of arrival of the direction-finding result as the center line, the distance estimated by the signal strength as the radius, and the error threshold as the half-angle. The time difference error band is defined as the area enclosed by two hyperbolas extending outwards from the TDOA observation value, with the curves not exceeding the positioning area. The TDOA-DPD method is used to construct a cost function in the pre-positioning region, divide the positioning grid, and determine the UAV position within the pre-positioning region using methods such as exhaustive search, gradient optimization, and genetic algorithms.

[0101] The pre-positioning region defined by AOA and TDOA dual-modal geometric constraints can greatly reduce the search mesh area of ​​TDOA-DPD, significantly reducing the computational load compared to the traditional TDOA-DPD method. Positioning is achieved through station angle measurement information, avoiding the fuzzy solution problem caused by TDOA. This solves the problems of excessive computational complexity and fuzzy single-modal positioning in existing technologies, as detailed below:

[0102] 1. Traditional TDOA-DPD solution can be abstracted in a physical sense as an exhaustive search over a possible location area to find the relative optimal solution of the maximum likelihood function. The method proposed in this embodiment of the invention reduces the amount of computation by reducing the location search area. It is simple to implement, applicable to multiple scenarios, and more in line with the actual production environment.

[0103] 2. This embodiment of the invention reduces the localization search area based on the multimodal geometric constraints of AOA and TDOA. Specifically, it constructs the intersection of the AOA angle measurement error cone and the TDOA time difference error band, which physically means using the angle measurement error cone to cut the TDOA time difference error band. Compared to the traditional DPD method, which must traverse the entire localization grid, the intersection area in this embodiment only occupies a small portion of the entire localization grid.

[0104] 3. The multimodal geometric constraint method provided in this embodiment of the invention constructs an angle measurement error cone and a time difference error band according to the actual electromagnetic environment, thereby forming different error regions; for example, the error cone sector region and error band region in a complex electromagnetic environment will be larger than the error region formed in a general electromagnetic environment, thereby avoiding the problem of increased error due to interference in the actual environment.

[0105] 4. Unlike general subspace-based DPD methods, the TDOA-DPD method used in this embodiment of the invention can make fuller use of the high-resolution characteristics of broadband signals in the time domain. By coherently accumulating the signals collected from all base stations, high-precision positioning with low signal-to-noise ratio is achieved, which is of great significance for non-cooperative positioning in complex electromagnetic environments in typical urban areas.

[0106] 5. The traditional TDOA-DPD method has an inherent TDOA positioning blind zone characteristic. The multimodal geometric constraint proposed in this embodiment of the invention determines the budget region through angle measurement information, which is not affected by the TDOA blind zone. It can solve the inherent blind zone problem of TDOA and does not affect the positioning accuracy of TDOA-DPD.

[0107] The above embodiments should be understood as being used only to illustrate the present invention more clearly, and not to limit the scope of the present invention. After reading the present invention, any modifications of the present embodiments by those skilled in the art will fall within the scope defined by the appended claims.

Claims

1. A direct localization method based on AOA and TDOA multimodal geometric constraints, characterized in that, include: Simultaneous acquisition of target signals emitted by radiation sources from multiple sites; The multi-site system includes multiple single sites, which are used to form a monitoring network deployed in the monitoring area; Obtain the AOA observation value of the single site, which is used to characterize the direction of arrival of the target signal; Obtain TDOA observations for base station pairs; each base station pair includes two single sites, one of which is a reference site and the other is the remaining single site excluding the reference site; An effective grid area is obtained based on the spatial intersection of the AOA angle measurement error cone and the TDOA time difference error band. The AOA angle measurement error cone is a fan-shaped area radiating outwards from the single station as the base point, the estimated distance from the radiation source to the single station as the radius, the direction of arrival as the center line, and the angle error threshold as half an angle. The TDOA time difference error band is the area enclosed by the extended curve. The extended curve is obtained through a branch curve and a distance error threshold. The branch curve is the target branch of the positioning hyperbola falling within the monitoring area. The positioning hyperbola is a hyperbola generated based on the distance difference with the base station pair as the focus. The distance difference is the distance difference from the radiation source to the base station pair, obtained by converting the TDOA observation values ​​of the base station pair. Based on the geometric positions of candidate target points within the effective grid area, the theoretical time difference of the target signal arriving at the base station pair is calculated, and the cross-correlation function value of the target signal received by the base station pair is calculated; effective cross-correlation function values ​​are selected based on the coherence coefficient threshold, and the effective cross-correlation function values ​​are accumulated to obtain the coherence accumulation result; Within the coordinate space corresponding to the effective grid region, the global maximum value of the coherent cumulative cost function is searched, and the coordinate point corresponding to the global maximum value is substituted into the direct localization algorithm to output the final estimated coordinates of the radiation source target; wherein, the coherent cumulative cost function is the global peak function of the coherent cumulative result corresponding to the coordinate point, and the coordinate point is a point within the coordinate space corresponding to the effective grid region.

2. The direct positioning method based on AOA and TDOA multimodal geometric constraints according to claim 1, characterized in that, The acquisition of the AOA observations specifically includes: The target signal is received by the array antenna of the single site, the target signal is processed by angle measurement, the frequency domain is transformed by discrete Fourier transform, and the AOA observation value representing the direction of arrival of the target signal is output by subspace analysis. The acquisition of the TDOA observations specifically includes: The base station synchronously receives the target signal emitted by the radiation source, converts it to the frequency domain through discrete Fourier transform, and obtains the frequency domain signal; based on the phase characteristics of the cross power spectrum of the frequency domain signal, the base station calculates the time delay difference between receiving the target signal and the target signal as the TDOA observation value.

3. The direct positioning method based on AOA and TDOA multimodal geometric constraints according to claim 1, characterized in that, The acquisition of the estimated distance from the radiation source to the single site specifically includes: Based on the received strength of the target signal, the estimated distance from the radiation source to the single site is obtained according to the spatial propagation loss model.

4. The direct positioning method based on AOA and TDOA multimodal geometric constraints according to claim 1, characterized in that, The generation of the effective grid region specifically includes: Construct a binarized sparse constraint matrix, which is the projection of a binarized mask function onto a discrete space; the discrete space is a set of candidate target points; the projection is the operation of mapping the output value of the binarized mask function at the candidate target points to the corresponding element of the binarized sparse constraint matrix. The binary mask function The range of values ​​is {0, 1}, and the mapping rule satisfies: When the candidate target point falls into the spatial intersection of the AOA angle measurement error cone and the TDOA time difference error band... ; When the candidate target point does not fall within the spatial intersection of the AOA angle measurement error cone and the TDOA time difference error band ; in, The coordinates of the candidate target point are given.

5. The direct positioning method based on AOA and TDOA multimodal geometric constraints according to claim 4, characterized in that, The specific formula for the relationship between the candidate target point falling into the AOA angle measurement error cone is as follows: ； in, For the first Location coordinates of a single site; For the first Angle of arrival obtained from actual measurements at a single site; The angle error threshold is... It is a positive integer. ; The number of the multiple sites; The specific formula for determining whether the candidate target point falls within the TDOA time difference error band is as follows: ； in, For the first Location coordinates of a single site; For the first Location coordinates of a single site; , , ; For the candidate target point obtained by actual measurement to reach the first Single station and arrival The time difference between individual sites, i.e., the time difference between the base stations TDOA observations; The speed of light; For the radiation source to the base station pair The distance difference, the first Each single site serves as a reference site; The distance error threshold is defined as follows.

6. The direct positioning method based on AOA and TDOA multimodal geometric constraints according to claim 5, characterized in that, The angle error threshold The distance error threshold ;in, The standard deviation of the azimuth angle. This represents the standard deviation of the time difference.

7. A direct positioning system based on AOA and TDOA multimodal geometric constraints, characterized in that, include: A broadband signal acquisition module is used to synchronously acquire target signals emitted by radiation sources from multiple sites. The multiple sites include multiple single sites, which are used to form a monitoring network deployed in the monitoring area. A multimodal parameter estimation module is used to obtain the AOA observation value of the single site, which is used to characterize the direction of arrival of the target signal; and to obtain the TDOA observation value of the base station pair; each base station pair includes two single sites, one of which is a reference site and the other is the remaining single site excluding the reference site. The effective grid area acquisition module is used to obtain an effective grid area based on the spatial intersection of the AOA angle measurement error cone and the TDOA time difference error band. The AOA angle measurement error cone is a fan-shaped area radiating outwards from the single station as the base point, the estimated distance from the radiation source to the single station as the radius, the direction of arrival as the center line, and the angle error threshold as half an angle. The TDOA time difference error band is the area enclosed by the extended curve. The extended curve is obtained through a branch curve and a distance error threshold. The branch curve is the target branch of the positioning hyperbola falling within the monitoring area. The positioning hyperbola is a hyperbola generated based on the distance difference with the base station pair as the focus. The distance difference is the distance difference from the radiation source to the base station pair, obtained by converting the TDOA observation values ​​of the base station pair. The coherent accumulation module is used to calculate the theoretical time difference of the target signal arriving at the base station pair based on the geometric position of the candidate target points in the effective grid area, and to calculate the cross-correlation function value of the target signal received by the base station pair; to filter out the effective cross-correlation function values ​​based on the coherence coefficient threshold, and to accumulate the effective cross-correlation function values ​​to obtain the coherent accumulation result; The target location determination module is used to search for the global maximum value of the coherent cumulative cost function in the coordinate space corresponding to the effective grid area, substitute the coordinate point corresponding to the global maximum value into the direct positioning algorithm, and output the final estimated coordinates of the radiation source target; wherein, the coherent cumulative cost function is the global peak function of the coherent cumulative result corresponding to the coordinate point, and the coordinate point is a point in the coordinate space corresponding to the effective grid area.

8. An electronic device comprising a memory, a processor, and a computer program or instructions stored in the memory, characterized in that, The processor executes the computer program or instructions to implement the direct positioning method based on AOA and TDOA multimodal geometric constraints as described in any one of claims 1 to 6.

Images