A radar networking layout simulation optimization system and method

Abstract

The application discloses a radar networking layout simulation optimization system and method, and relates to the technical field of radar networking layout simulation.The system comprises the following modules: a parameter acquisition module generates a candidate layout position set; a shielding judgment module judges whether a propagation path is shielded based on a building space model; a simulation calculation module calculates measurement accuracy under the condition of a non-shielded propagation path in combination with radar performance parameters, environment parameters and target parameters, and corrects signal amplitude and / or signal phase caused by reflection and / or scattering through a propagation calibration unit; a multi-radar fusion module generates confidence weight according to the measurement accuracy of each radar and fuses detection results to obtain networking detection accuracy; and a reverse simulation module optimizes search output radar layout scheme based on networking detection accuracy and networking performance indexes.Through the system and method, credible evaluation and optimized output of the layout scheme in a complex environment can be realized.

Description

Technical Field

[0001] This invention relates to the field of radar network deployment simulation technology, and in particular to a radar network deployment simulation optimization system and method. Background Technology

[0002] In the process of demonstrating and optimizing radar network deployment schemes, engineers typically need to comprehensively consider factors such as terrain, building obstruction, meteorological and electromagnetic environment, target characteristics, and radar performance parameters within a given working area. They must repeatedly simulate different combinations of candidate sites to obtain a network scheme that meets constraints such as coverage and accuracy. This type of demonstration is essentially a complex optimization problem with multiple coupled factors and strong constraints. It requires both reliable calculations of propagation conditions and measurement accuracy, and the evaluation and screening of a large number of candidate schemes within a limited time.

[0003] In existing technologies, common network deployment evaluation methods often employ ideal propagation assumptions or simplified propagation models, approximating propagation conditions using only geometric coverage or a single attenuation parameter, and then estimating measurement accuracy based on these approximations. Because signal amplitude and phase variations caused by factors such as building reflection and scattering are difficult to consistently handle in simplified models, and because there is a lack of a unified coupling mechanism between occlusion determination and accuracy calculation, the accuracy evaluation results of different candidate deployment schemes at the same target location are prone to deviation, making it difficult to accurately reflect the real detection conditions under complex urban scenarios and the combined effects of multiple environmental factors.

[0004] The above deviations will further amplify the risk of misjudgment in the optimization search process: on the one hand, the optimizer may misjudge candidate deployment locations with unreliable propagation conditions or insufficient measurement accuracy as preferred locations, and the output radar deployment scheme may not be able to stably meet the network performance indicators in the actual environment; on the other hand, faced with a large number of candidate deployment locations and schemes, existing technologies usually rely on manual experience or coarse-grained heuristic screening, which makes it difficult to ensure the credibility of the evaluation while taking into account the search efficiency, resulting in a longer scheme demonstration cycle, increased resource investment, and the risk of repeated adjustments and rework during the network construction phase. Summary of the Invention

[0005] To address the problem of how to reliably calculate the detection accuracy of a radar network under the coupled effects of urban buildings and meteorological / electromagnetic environment, and to automatically optimize radar deployment schemes accordingly, this invention proposes a radar network deployment simulation optimization system and method.

[0006] The present invention achieves the above objectives through the following technical solutions:

[0007] On one hand, the present invention provides a radar network deployment simulation and optimization system, the system comprising:

[0008] The parameter acquisition module is used to acquire the working area parameters, environmental parameters, target parameters and radar performance parameters required for radar network simulation, and generate a set of candidate deployment locations based on the working area parameters.

[0009] The obstruction determination module is used to determine whether there is obstruction in the propagation path between the radar and the target based on the building space model within the working area.

[0010] The simulation calculation module is used to calculate the radar's measurement accuracy at the candidate deployment location based on the radar performance parameters, environmental parameters, and target parameters, under the condition that the obstruction determination module determines the propagation path to be unobstructed. The simulation calculation module includes a propagation calibration unit, used to correct the signal amplitude and / or signal phase caused by reflection and / or scattering based on the propagation path characteristics of the radar signal along the propagation path, and to use the corrected signal amplitude and / or signal phase for calculating the corresponding measurement accuracy.

[0011] The multi-radar fusion module is used to generate corresponding confidence weights at the target location based on the measurement accuracy of each radar output by the simulation calculation module when the same target is detected by multiple radars located at different candidate deployment positions, and to fuse the multi-radar detection results according to the confidence weights to obtain network detection accuracy.

[0012] The reverse simulation module is used to optimize and search for candidate radar deployment schemes formed by the set of candidate deployment locations based on the network detection accuracy and according to preset network performance indicators, so as to output a radar deployment scheme that meets the network performance indicators.

[0013] A further improvement of the present invention is that the reverse simulation module includes:

[0014] The candidate space filtering unit is used to determine the contribution of each candidate deployment location to the network performance index based on the output results of the simulation calculation module and / or the multi-radar fusion module before the optimization search, and to filter the set of candidate deployment locations and / or assign different search weights according to the contribution. The search weights are used to adjust the selection probability of candidate deployment locations during the optimization search process.

[0015] The hierarchical evaluation unit is used to perform a first-stage performance evaluation on the candidate radar deployment schemes during the optimization search process. The first-stage performance evaluation determines whether the candidate radar deployment schemes meet preset conditions without calling the simulation calculation module. If the preset conditions are met, the simulation calculation module and the multi-radar fusion module are called to perform a second-stage simulation calculation on the candidate radar deployment schemes.

[0016] A further improvement of the present invention is that, for each candidate deployment location in the candidate deployment location set, the candidate space screening unit constructs a single radar deployment state that contains only that candidate deployment location, and based on the output results of the simulation calculation module and the multi-radar fusion module, determines the proportion of target locations in the target location set whose target locations satisfy the preset measurement accuracy threshold in the network performance index by the single radar deployment state, and determines the proportion of target locations as the contribution of the corresponding candidate deployment location to the network performance index;

[0017] Based on the stated contribution, perform at least one operation on the candidate deployment location set:

[0018] Candidate placement locations whose contribution is lower than the preset contribution threshold are removed from the candidate placement location set;

[0019] Candidate placement positions with a contribution level higher than or equal to a preset contribution level threshold are retained, and different search weights are assigned to the retained candidate placement positions according to the magnitude of the contribution level.

[0020] A further improvement of the present invention is that, when the hierarchical evaluation unit performs the first-stage performance evaluation of the candidate radar deployment scheme and triggers the second-stage simulation calculation during the optimization search process, it executes the following limited process:

[0021] The hierarchical evaluation unit obtains candidate radar deployment schemes consisting of multiple candidate deployment locations, and performs complementary coverage scoring on the candidate radar deployment schemes based on the contribution degree corresponding to each candidate deployment location output by the candidate space screening unit.

[0022] The complementary coverage score is calculated as follows:

[0023] Using the target location set as the evaluation object, the coverage of each candidate deployment location in the candidate radar deployment scheme to the target location set is judged one by one. When multiple candidate deployment locations cover the same target location, only the contribution of the candidate deployment location that achieves coverage for the first time is included in the score, thereby avoiding duplicate scoring of the same target location.

[0024] The tiered evaluation unit compares the complementary coverage score with a preset feasibility threshold.

[0025] When the complementary coverage score is lower than the preset feasibility threshold, the candidate radar deployment scheme is determined to not meet the preset conditions, and the subsequent evaluation of the candidate radar deployment scheme is terminated.

[0026] When the complementary coverage score reaches or exceeds the preset feasibility threshold, the hierarchical evaluation unit calls the simulation calculation module and the multi-radar fusion module to perform the second-stage simulation calculation on the candidate radar deployment scheme, and judges whether the candidate radar deployment scheme meets the network performance index based on the network detection accuracy output by the second-stage simulation calculation.

[0027] A further improvement of the present invention is that the occlusion determination module performs the following occlusion determination method on the propagation path between the radar and the target located at the candidate deployment position:

[0028] The electromagnetic material parameters of the building space model within the work area are obtained, as well as dynamic environmental element data reflecting the time-varying characteristics of the environment are extracted from the environmental parameters.

[0029] Based on the radar position and target position at the candidate deployment location, the direction of the propagation path between the radar and the target is determined, and the building surface that intersects the propagation path within a preset distance threshold is identified.

[0030] Based on the center frequency and polarization of the radar signal and the electromagnetic material parameters, calculate the signal diffraction loss and / or signal transmission loss caused by the building surface.

[0031] Based on the dynamic environmental element data, determine the additional signal attenuation caused by the dynamic environment along the propagation path;

[0032] The theoretical power of the radar transmitted signal is subtracted sequentially from the free space propagation loss of the propagation path, the signal diffraction loss and / or signal transmission loss, and the additional signal attenuation to obtain the estimated received signal power and signal margin at the target.

[0033] The minimum detectable signal power threshold is determined based on radar performance parameters, and the estimated received signal power is compared with the minimum detectable signal power threshold. When the estimated received signal power is greater than or equal to the minimum detectable signal power threshold, an unobstructed propagation path validity identifier is generated; otherwise, an electromagnetic obstruction propagation path validity identifier is generated.

[0034] The signal margin and propagation path effectiveness identifiers are output as quantifiable propagation condition information to the simulation calculation module.

[0035] A further improvement of the present invention is that the occlusion determination module further includes a propagation path state cache unit, which is used to store quantifiable propagation condition information indexed by radar position, target position, and environmental element identifiers determined by dynamic environmental element data.

[0036] Upon receiving an occlusion determination request, the propagation path cache unit is queried first, and if the query is successful, the cached quantifiable propagation condition information is directly output.

[0037] A further improvement of the present invention is that the propagation calibration unit employs a parameterized mapping function. Corrections are made to the signal amplitude and / or signal phase, specifically including:

[0038] The sequence of interaction points on the propagation path is determined based on the characteristics of the propagation path. The sequence of interaction points includes the incident angle and the material identifier of the building surface corresponding to the interaction point.

[0039] For each interaction point in the interaction point sequence, a feature vector is constructed using the center frequency of the radar signal, polarization mode, incident angle, and building surface material identifier. As a parameterized mapping function The input is used to obtain the magnitude correction amount at the corresponding interaction point. and phase correction , is represented as: ;in, Indicates the first One interaction point, For parameterized mapping functions The set of parameters;

[0040] Based on the total amplitude correction and the total phase correction, correction operations are performed on the signal amplitude and signal phase along the propagation path, satisfying: ;

[0041] In the formula, Indicates the number of interaction points; , These represent the signal amplitude and signal phase before correction, respectively. , These represent the corrected signal amplitude and signal phase, respectively.

[0042] A further improvement of the present invention is that, when the propagation path validity is identified as unobstructed, the simulation calculation module determines the equivalent noise power corresponding to the propagation path based on environmental parameters. and equivalent interference power And calculate the basic equivalent signal-to-noise ratio. The formula is: ;

[0043] In the formula, Signal margin; The target radar scattering characteristic parameter is one of the target parameters. Preset reference scattering characteristic parameters;

[0044] Using the corrected signal amplitude and the corrected signal phase Regarding the basic equivalent signal-to-noise ratio Perform combined amplitude and phase correction to obtain the final equivalent signal-to-noise ratio. The formula is: ;

[0045] In the formula, Indicates the preset reference signal amplitude. Indicates the preset reference signal phase;

[0046] Determine the accuracy of the reference distance measurement based on radar performance parameters. Accuracy of reference angle measurement and reference speed measurement accuracy And based on the final equivalent signal-to-noise ratio Calculate distance measurement accuracy Angle measurement accuracy and speed measurement accuracy The formula is: ;

[0047] When the propagation path validity is identified as having electromagnetic obstruction, the simulation calculation module sets the distance measurement accuracy, angle measurement accuracy, and velocity measurement accuracy to preset invalid values.

[0048] A further improvement of the present invention is that the multi-radar fusion module receives multiple sets of measurement accuracy results output by the simulation calculation module for the same target at the target location. Each set of measurement accuracy results corresponds to a radar located at the candidate deployment location and includes at least distance measurement accuracy, angle measurement accuracy, and velocity measurement accuracy.

[0049] For each radar, based on the geometric mapping relationship between range measurement accuracy and angle measurement accuracy at the target position, a target position-related uncertainty metric is constructed, satisfying: ;in For the first The uncertainty measure of the radar at the target location; Indicates the first The range measurement accuracy of the radar, Indicates the first The angle measurement accuracy of the radar. Indicates the first The speed measurement accuracy of the radar Indicates the first The distance between the radar and the target location. Preset weighting coefficients;

[0050] Initial confidence weights are generated based on the target position-related uncertainty measures corresponding to each radar. : ;in, This indicates the number of radars participating in fusion at the same target location. The first to participate in the integration Radar, For the first The uncertainty measure of the radar at the target location;

[0051] Based on the uncertainty metric distributions of multiple radars participating in the detection of the same target, suppression correction is performed on the initial confidence weights to generate the final confidence weights. : ;in, The mean of the uncertainty is... The inhibition coefficient;

[0052] Weights for final confidence level Normalization is performed, and weighted fusion is applied to the target measurement results from multiple radars:

[0053] ;in, Indicates the first Target measurement results corresponding to the radar This represents the target measurement results after multi-radar fusion;

[0054] Based on the fused target measurement results and the corresponding final confidence weights, the network detection accuracy at the target location is determined.

[0055] On the other hand, the present invention also provides a radar network deployment simulation optimization method, the method comprising:

[0056] The system acquires the working area parameters, environmental parameters, target parameters, and radar performance parameters required for radar network simulation, and generates a set of candidate deployment locations based on the working area parameters.

[0057] Based on the spatial model of buildings within the working area, determine whether there is any obstruction in the propagation path between the radar and the target;

[0058] Under the condition of an unobstructed propagation path, the measurement accuracy of the radar at the candidate deployment location is calculated based on the radar performance parameters, environmental parameters, and target parameters.

[0059] When the same target is detected by multiple radars located at different candidate deployment positions, a corresponding confidence weight is generated at the target position based on the measurement accuracy of each radar, and the detection results of multiple radars are fused according to the confidence weight to obtain the network detection accuracy.

[0060] Based on the network detection accuracy, and according to the preset network performance indicators, the candidate radar deployment scheme formed by the candidate deployment location set is optimized and searched to output a radar deployment scheme that meets the network performance indicators.

[0061] The beneficial effects of this invention are: by introducing the corrected signal amplitude and / or signal phase into the measurement accuracy calculation process, the propagation conditions and accuracy assessment form a consistent data flow and constraint relationship, thereby providing a stable and reusable basic input for subsequent network evaluation and optimization. It can integrate "propagation conditions—signal amplitude and phase—measurement accuracy—network fusion accuracy" in complex environments into the same simulation optimization link, reducing misjudgments caused by propagation simplification and evaluation separation, making the selection criteria for candidate deployment locations and candidate deployment schemes more consistent, and thus improving the reliability and engineering feasibility of network scheme demonstration and optimization outputs. Attached Figure Description

[0062] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0063] Figure 1 This is a schematic diagram of the modular structure of the system of the present invention;

[0064] Figure 2 This is a flowchart of the method in an embodiment of the present invention. Detailed Implementation

[0065] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention are within the scope of protection of the present invention.

[0066] like Figure 1 As shown, this is an embodiment of the present invention, which provides a radar network deployment simulation optimization system, including a parameter acquisition module, an obstruction determination module, a simulation calculation module, a multi-radar fusion module, and a reverse simulation module.

[0067] The parameter acquisition module is used to acquire the working area parameters, environmental parameters, target parameters and radar performance parameters required for radar network simulation, and to generate a set of candidate deployment locations based on the working area parameters.

[0068] The working area parameters are used to determine the task space range and sampling strategy, and include at least: the working area boundary (polygon endpoint coordinate sequence) and the simulation mesh resolution (mesh side length or sampling interval).

[0069] Environmental parameters are used to describe environmental conditions such as meteorology, electromagnetic interference, and background noise, and include at least: environmental noise-related parameters (such as ambient temperature, receiving bandwidth, noise figure, or directly giving the nominal value of noise power) and interference-related parameters (such as a list of interference sources or the nominal value of equivalent interference power).

[0070] The target parameters include at least the target radar scattering characteristics parameters (such as the target cross section or equivalent scattering index), and may include the target altitude range or target type identifier;

[0071] Radar performance parameters include at least: center frequency, polarization, theoretical power of transmitted signal, accuracy of reference distance measurement, accuracy of reference angle measurement, accuracy of reference velocity measurement, minimum detectable signal power threshold, or a set of parameters used to calculate the threshold.

[0072] The parameter acquisition module obtains working area parameters, environmental parameters, target parameters, and radar performance parameters through manual configuration files, databases, or external interfaces, and performs candidate deployment location set generation. The generation method includes at least one of the following:

[0073] Mesh generation method: Layout points are generated within the working area boundary according to the preset mesh spacing, and layout points in the prohibited area are removed;

[0074] Rule constraint method: On the basis of the grid generation method, height constraints and road / platform accessibility constraints are superimposed, and deployment points that meet the constraints are retained as candidate deployment locations.

[0075] The candidate deployment location set is stored in the form of a list or array. Each candidate deployment location contains at least spatial coordinates and deployment attribute identifiers (such as height layer and platform type).

[0076] The obstruction determination module is used to determine whether there is obstruction in the propagation path between the radar and the target based on the building space model within the working area.

[0077] In one preferred embodiment, the obstruction determination module, based on the building space model, electromagnetic material parameters, and dynamic environmental element data within the working area, makes a quantifiable determination of whether there is electromagnetic obstruction in the propagation path between the radar and the target at the candidate deployment location, and outputs the determination result to the simulation calculation module. A specific example is as follows:

[0078] The occlusion determination module obtains the following data in a single occlusion determination request:

[0079] Radar location: The three-dimensional coordinates (e.g., engineering coordinates or ENU coordinates) of the radar at the candidate deployment location in a unified coordinate system.

[0080] Target position: The three-dimensional coordinates of the target in the same coordinate system;

[0081] The building spatial model includes at least: a set of building surfaces (each building surface is represented by a polygonal or triangular patch, and each patch is defined by vertex coordinates), building surface normal vectors (used to calculate the angle of incidence), building surface material identifiers (used to associate electromagnetic material parameters), and patch bounding boxes (optionally used to accelerate spatial intersection retrieval).

[0082] Electromagnetic material parameters: These are configured using material identifiers as an index. They can be pre-configured from the engineering database or imported by the user from the environmental parameters. At a minimum, they include relative permittivity, conductivity, and material thickness (for transmission loss estimation).

[0083] Dynamic environmental element data: used to reflect the time-varying characteristics of the environment, which can come from external meteorological interfaces or internal environmental parameter configurations of the system. Before entering the occlusion determination module, the unit is unified and the time is aligned, including at least: timestamp, rainfall intensity (or other environmental elements that can be used for attenuation calculation).

[0084] Radar signal and radar performance parameters: radar signal center frequency, polarization mode (e.g., HH / VV / HV / VH, which can be represented by enumerated values), and theoretical power of radar transmitted signal; parameters in radar performance parameters used to determine the minimum detectable signal power threshold (the minimum detectable signal power threshold can be configured directly, or the noise figure, bandwidth and detection threshold can be configured to calculate the minimum detectable signal power threshold).

[0085] Based on the radar position and target position at the candidate deployment location, the direction of the propagation path between the radar and the target is determined, and the propagation path is represented as a line segment pointing from the radar position to the target position.

[0086] Set a preset distance threshold to accommodate discrete errors, coordinate errors, or mesh errors in the building model. The preset distance threshold can be 1–2 times the average side length of the building patch, or 0.5–1 times the side length of the simulation mesh cell.

[0087] The building surface recognition process employs a two-stage strategy: propagation path neighborhood retrieval + precise intersection judgment.

[0088] Level 1: Neighborhood Search

[0089] The propagation path is expanded into a path neighborhood with the propagation path as the axis, and the radius of the path neighborhood is set to a preset distance threshold. Building surfaces that spatially intersect with the path neighborhood are retrieved from the building spatial model to obtain a set of candidate building surfaces. In engineering implementation, it is recommended to use patch bounding boxes and hierarchical index structures (such as AABB trees / BVH) for retrieval: the bounding boxes corresponding to the propagation path neighborhood and the bounding boxes of the building surfaces are quickly eliminated, retaining only the candidate set.

[0090] Level 2: Precise Intersection Detection

[0091] For each facet in the candidate building surface set, determine whether the propagation path segment intersects the facet spatially. If the propagation path segment does not intersect the facet strictly, continue to calculate the shortest distance from the propagation path segment to the facet. If the shortest distance is not greater than a preset distance threshold, the facet is determined to have spatially intersected the propagation path within the preset distance threshold.

[0092] After completing the identification of candidate building surfaces, a set of intersecting surfaces of the propagation path is obtained, and the intersection position, surface normal vector, material identifier and incident angle are recorded for subsequent loss calculation.

[0093] Based on the radar signal center frequency, polarization, and electromagnetic material parameters, the signal diffraction loss and / or signal transmission loss caused by the building surface are calculated. To ensure feasibility, this embodiment provides a directly implementable method, and the system implementation also allows for substitution with an equivalent electromagnetic propagation model.

[0094] Signal diffraction loss calculation: When the propagation path forms a diffraction condition with the edge of a building, the position corresponding to the main obstruction edge on the propagation path is selected as the diffraction calculation point, and the diffraction loss is calculated using the single-edge diffraction approximation model. The required geometric quantities are obtained from the radar position, target position, and building surface geometry calculation; the polarization mode is used to select different coefficient tables or correction terms; if the propagation path does not form a diffraction condition, the diffraction loss is taken as 0.

[0095] Signal transmission loss calculation: When the propagation path passes through the building surface and enters the building medium, the transmission loss is calculated as interface transmission loss + medium absorption loss. Interface transmission loss is determined by the incident angle, polarization, and electromagnetic material parameters; medium absorption loss is determined by electromagnetic material parameters and material thickness. When the propagation path involves multiple penetrations, the transmission loss of each penetration is summed to obtain the total transmission loss. When the propagation path does not penetrate the building surface, the transmission loss is taken as 0.

[0096] The additional signal attenuation caused by the dynamic environment along the propagation path is determined based on dynamic environmental data. This embodiment uses rain attenuation as an example to illustrate the implementation: the rain attenuation model coefficients are determined based on the center frequency and polarization; the attenuation per unit length (dB / km) is calculated based on the rainfall intensity; the additional signal attenuation is obtained by multiplying the attenuation per unit length by the propagation path length. When the dynamic environmental data contains multiple types of elements, the total additional signal attenuation can be obtained by summing each element.

[0097] The free-space propagation loss is calculated based on the propagation path length and center frequency. The theoretical power of the radar transmitted signal is then subtracted sequentially from the free-space propagation loss, signal diffraction loss and / or signal transmission loss, and additional signal attenuation to obtain the estimated received signal power at the target location. The signal margin is then obtained from the difference between the estimated received signal power and the minimum detectable signal power threshold.

[0098] The minimum detectable signal power threshold is determined based on radar performance parameters, and the estimated received signal power is compared with the minimum detectable signal power threshold. When the estimated received signal power is greater than or equal to the minimum detectable signal power threshold, an unobstructed propagation path validity identifier is generated; otherwise, an electromagnetic obstruction propagation path validity identifier is generated.

[0099] The signal margin and propagation path validity identifier are encapsulated together into quantifiable propagation condition information and output to the simulation calculation module. When the simulation calculation module receives an unobstructed propagation path validity identifier, it continues to perform the measurement accuracy calculation for the target; if it receives a propagation path validity identifier with electromagnetic obstruction, it marks the corresponding measurement accuracy as invalid or sets it to a preset invalid value, thereby ensuring that the data source used by the subsequent multi-radar fusion module and the reverse simulation module is consistent and controllable.

[0100] Furthermore, the occlusion determination module also includes a propagation path state cache unit, used to reuse quantifiable propagation condition information to reduce redundant occlusion determination calculations. The propagation path state cache unit uses the following index key:

[0101] Radar position index: The radar position is quantized to discrete grid points according to a preset position quantization step size. The quantized discrete index is used as the index key.

[0102] Target location index: The target location is quantized to discrete grid points with the same quantization step size, and the quantized discrete index is used as the index key;

[0103] Environmental element identifier: generated from dynamic environmental element data. The generation method of environmental element identifier is as follows: the timestamp is discretized according to the preset time slice width to obtain the time slice number; the rainfall intensity is discretized into rainfall intensity levels according to the preset classification rules; the environmental element identifier is obtained by combining the time slice number and the rainfall intensity level.

[0104] The propagation path status cache unit uses a combination of "radar position index + target position index + environmental element identifier" as the cache index key, and the cache values ​​include: propagation path validity identifier, signal margin, estimated received signal power, and write timestamp.

[0105] When the occlusion determination module receives an occlusion determination request, it executes the following process:

[0106] Generate cache index keys based on radar location, target location, and dynamic environmental element data;

[0107] Look up the cache index key in the propagation path state cache unit:

[0108] When the query hits and the cache has not expired, the occlusion determination module directly outputs the quantifiable propagation condition information of the cache.

[0109] When a query fails or the cache expires, the occlusion determination module calls the aforementioned occlusion determination method to recalculate the quantifiable propagation condition information and writes the new result into the propagation path state cache unit.

[0110] The cache expiration rule can be set to a fixed validity period. It is recommended that the fixed validity period be consistent with or slightly less than the update cycle of dynamic environmental element data to ensure that the cache expires automatically when the environment changes.

[0111] The following is a recalculated example illustrating the calculation process and output results of the occlusion determination module. For simplicity, the example only considers rain attenuation as an additional signal attenuation due to the dynamic environment, and the diffraction loss and transmission loss are taken from the calculated values ​​of the model.

[0112] Example inputs include:

[0113] Center frequency: 10GHz;

[0114] Theoretical power of radar transmitted signal: 50dBm;

[0115] Distance between radar and target: 10km;

[0116] Dynamic environmental data: Rainfall intensity 25 mm / h;

[0117] Building electromagnetic material parameters: concrete material, the transmission loss calculation result corresponding to the penetration path is 15dB (including interface transmission and medium absorption).

[0118] Diffraction loss calculation result: 8dB;

[0119] Rain attenuation model: The rain attenuation coefficient per unit length is taken as 0.5 dB / km (corresponding to the coefficient table results under 25 mm / h, 10 GHz, and given polarization).

[0120] Minimum detectable signal power threshold: -105dBm (determined by radar performance parameters).

[0121] The calculation steps include:

[0122] Free-space propagation loss: The wavelength of 10 GHz is approximately 0.03 m. According to the standard formula, the free-space propagation loss is approximately 132.4 dB over a distance of 10 km.

[0123] Additional signal attenuation due to dynamic environment (rain attenuation): Total rain attenuation = 0.5dB / km × 10km = 5dB;

[0124] Total loss = Free space propagation loss 132.4dB + Diffraction loss 8dB + Transmission loss 15dB + Rain attenuation 5dB = 160.4dB;

[0125] Estimated received signal power = 50dBm - 160.4dB = -110.4dBm;

[0126] Signal margin = -110.4dBm - (-105dBm) = -5.4dB;

[0127] If the estimated received signal power is less than the minimum detectable signal power threshold, the obstruction determination module generates a propagation path validity identifier indicating the presence of electromagnetic obstruction.

[0128] Comparison example (no building diffraction / transmission, only rain attenuation): If the propagation path does not identify a building surface that intersects with the propagation path in space, the diffraction loss and transmission loss are taken as 0dB;

[0129] Total loss = 132.4dB + 5dB = 137.4dB;

[0130] Estimated received signal power = 50dBm - 137.4dB = -87.4dBm;

[0131] Signal margin = -87.4dBm - (-105dBm) = 17.6dB;

[0132] The occlusion determination module generates a validity identifier for an unobstructed propagation path and outputs corresponding quantifiable propagation condition information.

[0133] The simulation calculation module is used to calculate the radar's measurement accuracy at candidate deployment locations based on radar performance parameters, environmental parameters, and target parameters, under the condition that the obstruction determination module determines the propagation path to be unobstructed. It includes a propagation calibration unit, which corrects the signal amplitude and / or phase caused by reflection and / or scattering based on the propagation path characteristics of the radar signal, and uses the corrected signal amplitude and / or phase for calculating the corresponding measurement accuracy.

[0134] The propagation path features are provided by the occlusion determination module and the propagation path solving process, and at least include information on building surfaces that reflect and / or scatter along the propagation path. The propagation path features can be represented in data structure as a list of interaction events. Each interaction event includes: a building surface material identifier (e.g., an enumeration or string ID corresponding to "concrete," "glass," or "metal"), a building surface normal vector (used to calculate the angle of incidence), the incident direction vector of the propagation path at the interaction point (used to calculate the angle of incidence), and the spatial location of the interaction point (used for sorting and path consistency checks).

[0135] The propagation calibration unit determines the sequence of interaction points along the propagation path based on the propagation path characteristics. To avoid ambiguity, the interaction point sequence is ordered according to the propagation direction, with interaction points arranged sequentially from the radar position to the target position. The number of interaction points is denoted as [missing information]. .

[0136] For the first in the sequence of interaction points At each interaction point, the propagation calibration unit obtains the unit vector of the incident direction at that interaction point. and the corresponding building surface normal vector unit vector Calculate the angle of incidence This angle of incidence definition is applicable to both reflection and scattering scenarios and can be calculated directly. Each interaction point in the interaction point sequence includes: angle of incidence. Material markings on building surfaces.

[0137] Regarding the first Feature vectors are constructed from interaction points. To ensure the input is computable and has a fixed dimension, one possible implementation is:

[0138] Center frequency Input using normalized values;

[0139] The polarization method employs one-hot encoding or enumeration mapping to numerical encoding;

[0140] Angle of incidence Input in radians or angles;

[0141] Building surface material identification uses unique thermal coding or material index input;

[0142] eigenvectors Input parameterized mapping function The amplitude correction and phase correction are obtained and expressed as follows: ;in, Indicates the first The magnitude correction at each interaction point is defined as a positive real number; Indicates the first The phase correction at each interaction point is defined as a real number (in radians). This represents the set of parameters for a parameterized mapping function.

[0143] To ensure that those skilled in the art can implement it directly, parameterized mapping functions are provided. Any of the following implementation methods can be used:

[0144] Method A: Piecewise linear function or polynomial function (parameter set is coefficients);

[0145] Method B: Lookup table function (the parameter set is the table entry value for each material / angle / polarization / frequency combination);

[0146] Method C: Lightweight neural network function (parameter set is network weights).

[0147] The propagation calibration unit performs a deterministic combination of the interaction point outputs according to the interaction point sequence:

[0148] The total magnitude correction is calculated by multiplying: ;

[0149] The total phase correction is accumulated: .

[0150] Signal amplitude before correction Phase with the original signal This information originates from the radar received signal processing output. To ensure feasibility, a common implementation involves performing matched filtering and range-Doppler processing on the target echo to obtain the target echo complex envelope. ,Pick , If there are multiple observations in the system, the average complex envelope within the observation window can be used as the mean. .

[0151] For the signal amplitude before correction Phase with the original signal Perform the correction operation to satisfy: ;

[0152] In the formula, , These represent the corrected signal amplitude and signal phase, respectively.

[0153] The following gives the parameterized mapping function. An optional engineering implementation is used to output corresponding amplitude and phase corrections based on the feature vectors of the interaction points along the propagation path. In this embodiment, the parameterized mapping function... A lookup table model based on material and polarization grouping, combined with incident angle interpolation calculation, is employed to achieve this. This implementation does not rely on complex model training, the parameters have clear meanings, and it is suitable for radar simulation and engineering deployment scenarios.

[0154] The parameter set of the parameterized mapping function It consists of multiple correction parameter entries, each corresponding to a building surface material, polarization mode, and frequency band combination.

[0155] Regarding the first definition above Feature vectors of interaction points In this embodiment, the following processing is performed:

[0156] Material grouping: Maps the material identifiers of building surfaces to material category indexes, such as concrete, glass, metal, and composite materials;

[0157] Polarization grouping: Maps the radar signal polarization mode to a polarization index, such as HH, VV, HV, VH;

[0158] Frequency band grouping: grouping the center frequency of radar signals Map to a preset frequency band range, such as low frequency band, mid frequency band, and high frequency band;

[0159] Through the above steps, the feature vectors of the interaction points are... The mapping is to a specific parameter table index triplet: (material index, polarization index, frequency band index).

[0160] In the parameterized mapping function, a set of incident angle-correction mapping tables is pre-defined for each index triplet. This mapping table includes:

[0161] Several discrete incident angle nodes ;

[0162] Corresponding amplitude correction node ;

[0163] Corresponding phase correction node .

[0164] For the Each interaction point, the propagation calibration unit, is based on its incident angle. Interpolation calculations are performed in the corresponding mapping table to obtain the interaction point amplitude correction and interaction point phase correction. An optional implementation is linear interpolation.

[0165] like ,but:

[0166] ;

[0167] ;

[0168] When the incident angle exceeds the preset angle range, the propagation calibration unit can determine the correction amount by boundary value extrapolation or truncation.

[0169] In this embodiment, the parameter set of the parameterized mapping function It consists of the following:

[0170] The set of incident angle nodes corresponding to each material-polarization-frequency band combination;

[0171] The table entries for amplitude correction and phase correction for each incident angle node.

[0172] Parameter set It can be obtained or updated in any of the following ways:

[0173] Offline calibration method: Obtain correction scale items under different incident angle conditions through actual measurement or high-precision electromagnetic simulation;

[0174] Online update method: Incremental correction of table entries based on historical observation data.

[0175] It should be noted that this implementation is only an example of a parameterized mapping function. The parameterized mapping function can also be implemented in other equivalent ways, such as polynomial functions, piecewise functions, or neural network functions.

[0176] The simulation calculation module receives at least the following inputs when calculating measurement accuracy:

[0177] Propagation path validity identification and signal margin Quantifiable propagation condition information from the occlusion determination module. Signal margin. It is recommended to express this in dB;

[0178] Corrected signal amplitude Phase with the corrected signal : Output from the propagation calibration unit;

[0179] Environmental parameters: These must include at least data that can be used to determine the equivalent noise power and equivalent interference power at the receiver; for ease of implementation, an optional configuration includes: receiver bandwidth. Receiver noise figure Ambient temperature The list of interference sources or the nominal value of interference power can be imported from the parameter acquisition module or input from an external interface.

[0180] Target parameters: at least include target radar scattering characteristics. (For example, the target's scattering cross section or a scattering characteristic index equivalent to the scattering cross section), the target radar scattering characteristic parameters can be obtained from the target database, the target model calculation, or historical measurement data;

[0181] Radar performance parameters: including at least the accuracy of reference range measurement. Accuracy of reference angle measurement Accuracy of reference speed measurement The reference accuracy can be given by radar calibration data or performance manual.

[0182] When the propagation path validity is identified as unobstructed, the simulation calculation module determines the equivalent noise power corresponding to the propagation path based on environmental parameters. and equivalent interference power Optionally, it can be determined using power domain calculation.

[0183] Based on the receiving bandwidth in the environmental parameters Ambient temperature With receiver noise figure Calculate the thermal noise and convert it to receiver noise power: ,in This is the Boltzmann constant; if the nominal noise power value is directly given in the environmental parameters, the simulation calculation module directly uses the nominal value as... .

[0184] The equivalent interference power is calculated based on the interference source list in the environmental parameters. The interference source list can include the transmit power, frequency band occupancy, and equivalent coupling loss at the receiver for each interference source. The simulation calculation module calculates the equivalent interference power for each interference source at the receiver and sums them up to obtain the equivalent interference power. If the environmental parameters directly provide the nominal value of the equivalent interference power, the simulation calculation module will directly use the nominal value as... .

[0185] Preset reference scattering characteristic parameters The scattering correction factor can be calculated using either the nominal scattering characteristic parameters of a target of the same category or the scattering characteristic parameters of a standard calibration target. ;

[0186] Signal margin Equivalent noise power Equivalent interference power Scattering correction factor The combination yields the basic equivalent signal-to-noise ratio. The formula is: ;

[0187] Preset reference signal amplitude Phase with preset reference signal It can be obtained through any of the following methods:

[0188] Method A: The calibration echo complex envelope is calculated under standard calibration conditions;

[0189] Method B: Sampled and statistically analyzed typical unobstructed propagation paths during the system initialization phase.

[0190] Using the corrected signal amplitude and the corrected signal phase For the basic equivalent signal-to-noise ratio Perform combined amplitude and phase correction to obtain the final equivalent signal-to-noise ratio. The formula is: ;

[0191] Obtain the reference distance measurement accuracy based on radar performance parameters. Accuracy of reference angle measurement and reference speed measurement accuracy And based on the final equivalent signal-to-noise ratio Calculate distance measurement accuracy Angle measurement accuracy and speed measurement accuracy The formula is: ;

[0192] When the propagation path validity is identified as having electromagnetic obstruction, the simulation calculation module sets the distance measurement accuracy, angle measurement accuracy, and velocity measurement accuracy to preset invalid values.

[0193] The multi-radar fusion module is used to generate corresponding confidence weights at the target location based on the measurement accuracy of each radar output by the simulation calculation module when the same target is detected by multiple radars located at different candidate deployment positions. The multi-radar detection results are then fused according to the confidence weights to obtain the network detection accuracy.

[0194] The multi-radar fusion module receives at least the following input data:

[0195] Multiple sets of measurement accuracy results: The simulation calculation module outputs multiple sets of measurement accuracy results for the same target at the target location. Each set of measurement accuracy results corresponds to a radar located at the candidate deployment location and includes at least the range measurement accuracy. Angle measurement accuracy Speed ​​measurement accuracy ;in, This represents the radar index, and the number of radars participating in the fusion is... The radar index satisfies ;

[0196] Target measurement results from multiple radars: Each radar outputs a target measurement result vector at the same target location. In one optional implementation, the target measurement result vector takes the following form: ,in This is a distance measurement value. This is an angle measurement value. This is a speed measurement value;

[0197] Radar position and target position: The multi-radar fusion module receives the radar position corresponding to each radar. With the target location Used to calculate the geometric distance between the radar and the target location. : .

[0198] For each radar, based on the geometric mapping relationship between range measurement accuracy and angle measurement accuracy at the target position, a target position-related uncertainty metric is constructed, satisfying: ;in For the first The uncertainty measure of the radar at the target location; Indicates the first The distance between the radar and the target location. These are preset weighting coefficients used to adjust the contribution of different measurement dimensions to the uncertainty measure;

[0199] To facilitate engineering implementation, the preset weighting coefficients can be determined using one of the following methods:

[0200] Fixed configuration method: Set it as a constant and write it to the system configuration file;

[0201] Indicator alignment method: Different weighting coefficients are set according to the degree of attention paid to distance / angle / speed accuracy of network performance indicators.

[0202] Using reciprocal normalization, the uncertainty measure Mapped to initial confidence weights : ;in, The first to participate in the integration Radar, For the first The uncertainty measure of the radar at the target location; satisfy This calculation method ensures that radars with smaller uncertainty correspond to larger initial confidence weights, thus establishing a deterministic correspondence between weights and measurement accuracy.

[0203] To avoid distortion of the initial weights when outliers exist in the uncertain metric distribution, a suppression correction is performed on the initial confidence weights based on the uncertain metric distribution to generate the final confidence weights. : ;in, The mean of the uncertainty is... ; It is the suppression coefficient and is a non-negative real number;

[0204] To ensure the weight sum is 1, the multi-radar fusion module performs normalization processing on the final confidence weights, making... After normalization, it satisfies .

[0205] Using final confidence weights Target measurement results from multiple radars Execute weighted fusion: ;in, Indicates the first Target measurement results corresponding to the radar This represents the target measurement results after multi-radar fusion;

[0206] Based on the fused target measurement results and the corresponding final confidence weights, the network detection accuracy at the target location is determined.

[0207] Multi-radar fusion module output:

[0208] Fusion target measurement results ;

[0209] Network detection accuracy at the target location.

[0210] In one optional implementation, the network detection accuracy is represented by a comprehensive accuracy index of fused range, angle, and velocity. Specifically:

[0211] The multi-radar fusion module combines the measurement accuracies of each radar according to the final confidence weight to obtain the range, angle, and velocity components of the network detection accuracy, satisfying the following: ;

[0212] in, , , These represent the network detection accuracy components at the target location.

[0213] The reverse simulation module is used to optimize and search for candidate radar deployment schemes formed by a set of candidate deployment locations, based on network detection accuracy and preset network performance indicators, in order to output a radar deployment scheme that meets the network performance indicators. The network performance indicators include at least coverage requirements and measurement accuracy requirements, and optionally include redundancy requirements.

[0214] In one preferred embodiment, the reverse simulation module includes a candidate space screening unit and a hierarchical evaluation unit.

[0215] Candidate radar deployment schemes are represented as a set or ordered list of candidate deployment locations. Each element in a candidate radar deployment scheme is a candidate deployment location in the set of candidate deployment locations, and may further include constraints on the number of radars, different radar types, or deployment spacing. The form of the constraints is consistent with the network performance indicators or engineering constraints.

[0216] The reverse simulation module employs an iterative optimization search. Each iteration generates one or more candidate radar deployment schemes and calls the hierarchical evaluation unit to perform the first-stage performance evaluation and the second-stage simulation calculation.

[0217] During the generation of candidate radar deployment schemes, the reverse simulation module uses the search weights output by the candidate space filtering unit to adjust the selection probability of candidate deployment locations. One possible approach is to perform weighted random sampling in the candidate deployment location set according to the search weights, with the number of samplings equal to the number of radars in the candidate radar deployment scheme. During the sampling process, a constraint of "the same candidate deployment location cannot be selected repeatedly" can be added.

[0218] The hierarchical evaluation unit first performs a first-stage performance evaluation for each candidate radar deployment scheme. The first-stage performance evaluation does not call the simulation calculation module; it only uses the contribution and coverage information output by the candidate space screening unit to calculate the score and compare it with the feasibility threshold.

[0219] When the first-stage performance evaluation meets the preset conditions, the hierarchical evaluation unit calls the simulation calculation module and the multi-radar fusion module to perform the second-stage simulation calculation on the candidate radar deployment scheme, obtains the network detection accuracy at the target location, and compares it with the network performance indicators.

[0220] Specifically, the contribution and search weight generation of the candidate space screening units include:

[0221] The candidate space screening unit establishes a set of target locations. This set can come from grid points in the mission area, target trajectory sampling points, or sets of key protection points. The preset measurement accuracy thresholds in the network performance indicators are determined by the system configuration. These thresholds can include distance measurement accuracy thresholds, angle measurement accuracy thresholds, and velocity measurement accuracy thresholds, with the threshold format consistent with the actual evaluation criteria.

[0222] The candidate space filtering unit constructs a single radar deployment state for each candidate deployment location in the candidate deployment location set. A single radar deployment state indicates that only one radar is deployed at that candidate deployment location. For each single radar deployment state, the candidate space filtering unit performs a coverage determination for each target location in the target location set. The coverage determination includes at least the following steps:

[0223] (1) Call the obstruction determination module to determine whether the propagation path between the radar and the target position is unobstructed under the single radar deployment state;

[0224] (2) Under unobstructed conditions, the simulation calculation module is called to obtain the distance measurement accuracy, angle measurement accuracy and velocity measurement accuracy of a single radar to the target position;

[0225] (3) Compare the distance measurement accuracy, angle measurement accuracy and speed measurement accuracy with the preset measurement accuracy threshold. When all three accuracies meet the threshold constraints, mark the target position as "covered by the candidate deployment position".

[0226] To facilitate subsequent complementary coverage score calculation, the candidate space filtering unit saves the coverage result of each candidate deployment location on the target location set as a Boolean array or bitmap. Each bit of the Boolean array corresponds to a target location, and the value indicates whether it is covered or not.

[0227] The candidate space screening unit counts the number of target locations covered by a single radar deployment status and calculates the proportion of the covered target locations to the total number of target locations. This proportion is determined as the contribution of the corresponding candidate deployment location to the network performance index. Based on the contribution, at least one operation is performed on the candidate deployment location set:

[0228] Candidate placement locations with a contribution value lower than a preset contribution value threshold are removed from the candidate placement location set;

[0229] Candidate placement positions with a contribution level higher than or equal to a preset contribution threshold will be retained.

[0230] The retained candidate placement positions are assigned different search weights based on their contribution. One possible implementation is as follows:

[0231] Sort the contributions from largest to smallest;

[0232] The candidate placement positions that rank higher are assigned higher search weights, while the candidate placement positions that rank lower are assigned lower search weights.

[0233] Perform normalization on all search weights so that the sum of the search weights is 1.

[0234] The candidate space filtering unit outputs the following data for use by the hierarchical evaluation unit and the optimization search:

[0235] The contribution of each candidate deployment location;

[0236] A Boolean array or bitmap covering each candidate placement location;

[0237] The search weight of each candidate deployment location;

[0238] The set of candidate deployment locations after elimination.

[0239] The complementary coverage scoring and grading triggers of the tiered assessment units include:

[0240] The hierarchical evaluation unit receives candidate radar deployment schemes, which include multiple candidate deployment locations. The hierarchical evaluation unit further receives a Boolean array or bitmap of contribution and coverage output by the candidate spatial filtering unit.

[0241] The hierarchical evaluation unit uses the target location set as the evaluation object and performs complementary coverage scoring. To ensure that the logic of "scoring for first-time coverage and not scoring for repeated coverage" can be directly implemented, the hierarchical evaluation unit adopts the following optional implementation methods:

[0242] (1) Initialize the “uncovered target location set” to all target locations in the target location set;

[0243] (2) Sort the candidate radar deployment locations in the candidate radar deployment schemes from largest to smallest according to their contribution, and obtain a sorted list;

[0244] (3) Traverse each candidate placement position in the sorted list order:

[0245] Read the Boolean array or bitmap corresponding to the candidate deployment positions;

[0246] Calculate the "new set of covered target locations". The new set of covered target locations is the target locations marked as covered in the Boolean array or bitmap and belonging to the set of uncovered target locations.

[0247] When the set of newly added coverage target locations is not empty, the contribution of the candidate deployment locations is included in the complementary coverage score, and the set of newly added coverage target locations is deleted from the set of uncovered target locations.

[0248] When the set of newly added target locations is empty, no contribution is added to the complementary coverage score, thereby avoiding duplicate scoring of the same target location.

[0249] The above process ensures that when multiple candidate deployment locations cover the same target location, only the contribution of the candidate deployment location that achieves coverage for the first time is included in the scoring.

[0250] The tiered assessment unit compares the complementary coverage score with a preset feasibility threshold:

[0251] When the complementary coverage score is lower than the preset feasibility threshold, the candidate radar deployment scheme is determined not to meet the preset conditions, the subsequent evaluation of the candidate radar deployment scheme is terminated, and the candidate radar deployment scheme is marked as infeasible.

[0252] When the complementary coverage score reaches or exceeds the preset feasibility threshold, the candidate radar deployment scheme is determined to meet the preset conditions of the first stage, and enters the second stage simulation calculation.

[0253] The preset feasibility threshold can be determined by system configuration. One possible implementation is to set a lower limit ratio of the feasibility threshold based on the radar quantity constraint and the size of the target location set, so that the first-stage evaluation can eliminate candidate radar deployment schemes with obvious insufficient coverage.

[0254] After the candidate radar deployment scheme passes the first-stage performance evaluation, the hierarchical evaluation unit calls the simulation calculation module and the multi-radar fusion module to perform the second-stage simulation calculation. The second-stage simulation calculation includes at least the following:

[0255] (1) Perform occlusion determination and simulation calculation on each candidate deployment location and target location set in the candidate radar deployment scheme to obtain the measurement accuracy of each radar at the target location;

[0256] (2) Call the multi-radar fusion module to generate confidence weights at the target location and fuse the detection results of multiple radars to obtain the network detection accuracy at the target location;

[0257] (3) Compare the network detection accuracy at the target location with the network performance index. If the network performance index is met, output the candidate radar deployment scheme as the radar deployment scheme that meets the network performance index. If it is not met, determine the candidate radar deployment scheme as not meeting the network performance index and return to the optimization search to continue iterating.

[0258] like Figure 2 As shown, another embodiment of the present invention provides a radar network deployment simulation optimization method, including the following steps:

[0259] The system acquires the working area parameters, environmental parameters, target parameters, and radar performance parameters required for radar network simulation, and generates a set of candidate deployment locations based on the working area parameters.

[0260] Based on the spatial model of buildings within the working area, determine whether there is any obstruction in the propagation path between the radar and the target;

[0261] Under the condition of an unobstructed propagation path, the measurement accuracy of the radar at the candidate deployment location is calculated based on radar performance parameters, environmental parameters, and target parameters.

[0262] When the same target is detected by multiple radars located at different candidate deployment positions, a corresponding confidence weight is generated at the target position based on the measurement accuracy of each radar, and the detection results of multiple radars are fused according to the confidence weight to obtain the network detection accuracy.

[0263] Based on the network detection accuracy, and according to the preset network performance indicators, the candidate radar deployment scheme formed by the candidate deployment location set is optimized and searched to output a radar deployment scheme that meets the network performance indicators.

[0264] In summary, this invention forms a closed-loop evaluation link for network detection accuracy by determining propagation path obstruction based on building spatial models, calibrating signal amplitude / phase based on propagation path characteristics, calculating measurement accuracy by combining radar performance parameters / environmental parameters / target parameters, and multi-radar fusion based on measurement accuracy to generate confidence weights. The reverse simulation module completes the optimization search of candidate deployment schemes under network performance index constraints, thereby improving the consistency of deployment scheme evaluation and the usability of scheme output in complex scenarios.

[0265] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various variations or substitutions within the technical scope disclosed in this application, and these should all be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A radar network deployment simulation and optimization system, characterized in that, The system includes: The parameter acquisition module is used to acquire the working area parameters, environmental parameters, target parameters and radar performance parameters required for radar network simulation, and generate a set of candidate deployment locations based on the working area parameters. The obstruction determination module is used to determine whether there is obstruction in the propagation path between the radar and the target based on the building space model within the working area. The simulation calculation module is used to calculate the radar's measurement accuracy at the candidate deployment location based on the radar performance parameters, environmental parameters, and target parameters, under the condition that the obstruction determination module determines the propagation path to be unobstructed. The simulation calculation module includes a propagation calibration unit, used to correct the signal amplitude and phase caused by reflection and / or scattering based on the propagation path characteristics of the radar signal along the propagation path, and to use the corrected signal amplitude and phase for calculating the corresponding measurement accuracy. The multi-radar fusion module is used to generate corresponding confidence weights at the target location based on the measurement accuracy of each radar output by the simulation calculation module when the same target is detected by multiple radars located at different candidate deployment positions, and to fuse the multi-radar detection results according to the confidence weights to obtain network detection accuracy. The reverse simulation module is used to optimize and search for candidate radar deployment schemes formed by the set of candidate deployment locations based on the network detection accuracy and according to preset network performance indicators, so as to output a radar deployment scheme that meets the network performance indicators.

2. The radar network deployment simulation and optimization system according to claim 1, characterized in that, The reverse simulation module includes: The candidate space filtering unit is used to determine the contribution of each candidate deployment location to the network performance index based on the output results of the simulation calculation module before the optimization search, and to filter the set of candidate deployment locations and / or assign different search weights according to the contribution. The search weights are used to adjust the selection probability of candidate deployment locations during the optimization search process. The hierarchical evaluation unit is used to perform a first-stage performance evaluation on the candidate radar deployment schemes during the optimization search process. The first-stage performance evaluation determines whether the candidate radar deployment schemes meet preset conditions without calling the simulation calculation module. If the preset conditions are met, the simulation calculation module and the multi-radar fusion module are called to perform a second-stage simulation calculation on the candidate radar deployment schemes.

3. The radar network deployment simulation and optimization system according to claim 2, characterized in that, For each candidate deployment location in the candidate deployment location set, the candidate space screening unit constructs a single radar deployment state that contains only that candidate deployment location. Based on the output of the simulation calculation module, it determines the proportion of target locations in the target location set that satisfy the preset measurement accuracy threshold in the network performance index by the single radar deployment state. The proportion of target locations is determined as the contribution of the corresponding candidate deployment location to the network performance index. Based on the stated contribution, perform at least one operation on the candidate deployment location set: Candidate placement locations whose contribution is lower than the preset contribution threshold are removed from the candidate placement location set; Candidate placement positions with a contribution level higher than or equal to a preset contribution level threshold are retained, and different search weights are assigned to the retained candidate placement positions according to the magnitude of the contribution level.

4. The radar network deployment simulation and optimization system according to claim 3, characterized in that, When the hierarchical evaluation unit performs the first-stage performance evaluation of candidate radar deployment schemes and triggers the second-stage simulation calculation during the optimization search process, it executes the following limited process: The hierarchical evaluation unit obtains candidate radar deployment schemes consisting of multiple candidate deployment locations, and performs complementary coverage scoring on the candidate radar deployment schemes based on the contribution degree corresponding to each candidate deployment location output by the candidate space screening unit. The complementary coverage score is calculated as follows: Using the target location set as the evaluation object, the coverage of each candidate deployment location in the candidate radar deployment scheme to the target location set is judged one by one. When multiple candidate deployment locations cover the same target location, only the contribution of the candidate deployment location that achieves coverage for the first time is included in the score, thereby avoiding duplicate scoring of the same target location. The tiered evaluation unit compares the complementary coverage score with a preset feasibility threshold. When the complementary coverage score is lower than the preset feasibility threshold, the candidate radar deployment scheme is determined to not meet the preset conditions, and the subsequent evaluation of the candidate radar deployment scheme is terminated. When the complementary coverage score reaches or exceeds the preset feasibility threshold, the hierarchical evaluation unit calls the simulation calculation module and the multi-radar fusion module to perform the second-stage simulation calculation on the candidate radar deployment scheme, and judges whether the candidate radar deployment scheme meets the network performance index based on the network detection accuracy output by the second-stage simulation calculation.

5. The radar network deployment simulation and optimization system according to claim 1, characterized in that, The obstruction determination module performs the following obstruction determination method on the propagation path between the radar and the target located at the candidate deployment position: The electromagnetic material parameters of the building space model within the work area are obtained, as well as dynamic environmental element data reflecting the time-varying characteristics of the environment are extracted from the environmental parameters. Based on the radar position and target position at the candidate deployment location, the direction of the propagation path between the radar and the target is determined, and the building surface that intersects the propagation path within a preset distance threshold is identified. Based on the center frequency and polarization of the radar signal and the electromagnetic material parameters, calculate the signal diffraction loss and / or signal transmission loss caused by the building surface. Based on the dynamic environmental element data, determine the additional signal attenuation caused by the dynamic environment along the propagation path; The theoretical power of the radar transmitted signal is subtracted sequentially from the free space propagation loss of the propagation path, the signal diffraction loss and / or signal transmission loss, and the additional signal attenuation to obtain the estimated received signal power and signal margin at the target. The minimum detectable signal power threshold is determined based on radar performance parameters, and the estimated received signal power is compared with the minimum detectable signal power threshold. When the estimated received signal power is greater than or equal to the minimum detectable signal power threshold, an unobstructed propagation path validity identifier is generated; otherwise, an electromagnetic obstruction propagation path validity identifier is generated. The signal margin and propagation path effectiveness identifiers are output as quantifiable propagation condition information to the simulation calculation module.

6. The radar network deployment simulation and optimization system according to claim 5, characterized in that, The occlusion determination module also includes a propagation path state cache unit, which stores quantifiable propagation condition information indexed by radar position, target position, and environmental element identifiers determined by dynamic environmental element data. Upon receiving an occlusion determination request, the propagation path cache unit is queried first, and if the query is successful, the cached quantifiable propagation condition information is directly output.

7. The radar network deployment simulation and optimization system according to claim 6, characterized in that, The propagation calibration unit employs a parameterized mapping function. Corrections are made to the signal amplitude and phase, specifically including: The sequence of interaction points on the propagation path is determined based on the characteristics of the propagation path. The sequence of interaction points includes the incident angle and the material identifier of the building surface corresponding to the interaction point. For each interaction point in the interaction point sequence, a feature vector is constructed using the center frequency of the radar signal, polarization mode, incident angle, and building surface material identifier. As a parameterized mapping function The input is used to obtain the magnitude correction amount at the corresponding interaction point. and phase correction , is represented as: ;in, Indicates the first One interaction point, For parameterized mapping functions The set of parameters; Based on the total amplitude correction and the total phase correction, correction operations are performed on the signal amplitude and signal phase along the propagation path, satisfying: ; In the formula, Indicates the number of interaction points; , These represent the signal amplitude and signal phase before correction, respectively. , These represent the corrected signal amplitude and signal phase, respectively.

8. The radar network deployment simulation and optimization system according to claim 7, characterized in that, When the propagation path validity is identified as unobstructed, the simulation calculation module determines the equivalent noise power corresponding to the propagation path based on environmental parameters. and equivalent interference power And calculate the basic equivalent signal-to-noise ratio. The formula is: ; In the formula, Signal margin; The target radar scattering characteristic parameter is one of the target parameters. Preset reference scattering characteristic parameters; Using the corrected signal amplitude and the corrected signal phase Regarding the basic equivalent signal-to-noise ratio Perform combined amplitude and phase correction to obtain the final equivalent signal-to-noise ratio. The formula is: ; In the formula, Indicates the preset reference signal amplitude. Indicates the preset reference signal phase; Determine the accuracy of the reference distance measurement based on radar performance parameters. Accuracy of reference angle measurement and reference speed measurement accuracy And based on the final equivalent signal-to-noise ratio Calculate distance measurement accuracy Angle measurement accuracy and speed measurement accuracy The formula is: ; When the propagation path validity is identified as having electromagnetic obstruction, the simulation calculation module sets the distance measurement accuracy, angle measurement accuracy, and velocity measurement accuracy to preset invalid values.

9. The radar network deployment simulation and optimization system according to claim 8, characterized in that, The multi-radar fusion module receives multiple sets of measurement accuracy results output by the simulation calculation module for the same target at the target location. Each set of measurement accuracy results corresponds to a radar located at the candidate deployment location and includes at least distance measurement accuracy, angle measurement accuracy, and velocity measurement accuracy. For each radar, based on the geometric mapping relationship between range measurement accuracy and angle measurement accuracy at the target position, a target position-related uncertainty metric is constructed, satisfying: ;in For the first The uncertainty measure of the radar at the target location; Indicates the first The range measurement accuracy of the radar, Indicates the first The angle measurement accuracy of the radar. Indicates the first The speed measurement accuracy of the radar Indicates the first The distance between the radar and the target location. Preset weighting coefficients; Initial confidence weights are generated based on the target position-related uncertainty measures corresponding to each radar. : ;in, This indicates the number of radars participating in fusion at the same target location. The first to participate in the integration Radar, For the first The uncertainty measure of the radar at the target location; Based on the uncertainty metric distributions of multiple radars participating in the detection of the same target, suppression correction is performed on the initial confidence weights to generate the final confidence weights. : ;in, The mean of the uncertainty, The inhibition coefficient; Weights for final confidence level Normalization is performed, and weighted fusion is applied to the target measurement results from multiple radars: ;in, Indicates the first Target measurement results corresponding to the radar This represents the target measurement results after multi-radar fusion; Based on the fused target measurement results and the corresponding final confidence weights, the network detection accuracy at the target location is determined.

10. A radar network deployment simulation optimization method, based on the radar network deployment simulation optimization system according to any one of claims 1-9, characterized in that, The method includes: The system acquires the working area parameters, environmental parameters, target parameters, and radar performance parameters required for radar network simulation, and generates a set of candidate deployment locations based on the working area parameters. Based on the spatial model of buildings within the working area, determine whether there is any obstruction in the propagation path between the radar and the target; Under the condition of an unobstructed propagation path, the measurement accuracy of the radar at the candidate deployment location is calculated based on the radar performance parameters, environmental parameters, and target parameters. When the same target is detected by multiple radars located at different candidate deployment positions, a corresponding confidence weight is generated at the target position based on the measurement accuracy of each radar, and the detection results of multiple radars are fused according to the confidence weight to obtain the network detection accuracy. Based on the network detection accuracy, and according to the preset network performance indicators, the candidate radar deployment scheme formed by the candidate deployment location set is optimized and searched to output a radar deployment scheme that meets the network performance indicators.

Images