Search

CN-122002214-A - General sense integrated physical environment reconstruction method

CN122002214ACN 122002214 ACN122002214 ACN 122002214ACN-122002214-A

Abstract

The invention belongs to the technical field of wireless communication and positioning, and particularly relates to a general sense integrated physical environment reconstruction method. The method comprises the steps of taking electromagnetic wave propagation characteristics as an information carrier, capturing multidimensional information such as environmental physical parameters and the like in real time through a multi-mode sensing means, extracting key environmental characteristics through a multi-source information fusion algorithm, constructing a parameterized digital twin model, calculating, evaluating and adjusting the electromagnetic wave propagation characteristics, adopting an optimized ultra-fast electric field calculation method, combining an environment-channel correlation database, deducing electromagnetic propagation behaviors in real time in a reconstruction scene, outputting path parameters, forming a closed-loop autonomous system of perception-prediction-behaviors, converting a prediction result into an optimization strategy, reversely adjusting the perception and prediction parameters, and realizing overall and local two-stage closed-loop feedback. The method solves the problems of large ranging error, large channel prediction deviation and high deployment cost by multiplexing the OFDM waveform, multi-mode sensing fusion and carrier phase ranging.

Inventors

  • ZHAO RUNZE
  • ZHAO LI
  • HAN SHUNLI
  • ZHANG KUI
  • WANG TONG

Assignees

  • 中国电子科技集团公司第四十一研究所

Dates

Publication Date
20260508
Application Date
20260105

Claims (10)

  1. 1. The utility model provides a general sense integrated physical environment reconstruction method which is characterized in that electromagnetic wave propagation characteristics are used as an information carrier, and the method comprises the following steps: s1, environment sensing, namely capturing multidimensional information of environment physical parameters, obstacles and mobile scatterers in real time through a multi-mode sensing means; S2, constructing a virtual physical environment, extracting key environment characteristics through a multi-source information fusion algorithm, and constructing a parameterized digital twin model, so that the shape, the position, the materials and the motion state of each object in the virtual physical environment are mapped with the real world in a high precision; S3, predicting electromagnetic wave propagation, calculating, evaluating and adjusting the electromagnetic wave propagation characteristics, adopting an optimized ultra-fast electric field calculation method, combining an environment-channel association database, and deducing electromagnetic propagation behaviors in real time in a reconstruction scene to output path parameters, wherein the path parameters comprise but are not limited to propagation directions, amplitudes, delays and phases; s4, forming a closed-loop autonomous system of perception-prediction-behavior, converting a prediction result into an optimization strategy, and reversely adjusting perception and prediction parameters to realize global and local two-stage closed-loop feedback.
  2. 2. The general sense integrated physical environment reconstruction method according to claim 1, wherein the physical environment reconstruction method is realized based on a wireless environment prediction and reconstruction system, and the system adopts a five-layer architecture, namely an infrastructure layer, a resource layer, a network function layer, an application layer and a prediction plane crossing all layers, wherein an environment sensing module, an electromagnetic wave propagation prediction module and a behavior module are embedded in the prediction plane to form two-stage closed loop feedback.
  3. 3. The method for reconstructing a physical environment with integrated sense of general according to claim 2, wherein, The first-stage closed-loop feedback penetrates through the physical environment full layer, wherein an original perception sample is provided by a basic facility and a resource layer, a perception module extracts knowledge and reconstructs a virtual environment, a prediction module calculates and evaluates electromagnetic propagation characteristics, a behavior module generates an optimization strategy and feeds the optimization strategy back to a network function layer and an application layer, and global adaptation is realized; The second-stage closed-loop feedback exists between adjacent modules, namely, the sensing module and the prediction module realize quick fine adjustment through a model refreshing signal, and the prediction module and the behavior module realize instant response through the requirement-strategy conversion so as to carry out local fine adjustment.
  4. 4. The method for reconstructing a physical environment with integrated sense of general according to claim 3, wherein the electromagnetic wave propagation prediction module adopts a ranging method based on Orthogonal Frequency Division Multiplexing (OFDM) time delay, comprising the steps of: S301, symbol coarse synchronization and pilot frequency detection, namely directly representing the arrival time of an integer sampling stage of a direct path by using a peak position by adopting a delay correlation algorithm based on cyclic prefix; S302, extracting effective multipath information from a pilot sequence by a multi-frequency multipath capturing method, extracting the initial path information from the effective multipath, and providing an initial value for subsequent fine synchronization; s303, synchronizing multipath tracking with the coarse symbol, correcting the head path position through a differential channel power balance time delay estimation algorithm to obtain high-precision time delay estimation, combining a carrier phase ranging algorithm, projecting a received pilot signal onto an ideal head path reference signal generated based on the high-precision time delay estimation, and extracting the phase of a projection coefficient as a carrier phase observation value.
  5. 5. The method for reconstructing a physical environment with integrated sense of general purpose as claimed in claim 4, wherein in step S302, a capturing window is formed by intercepting a small range of sampling points with a delay estimated value obtained by coarse synchronization as a center; Constructing a frequency domain phase rotation vector for each candidate time delay in a capturing window, calculating the projection power of a received pilot frequency on the vector, traversing all candidate time delays, selecting a plurality of paths with strongest power as effective multipaths, analyzing the effective multipaths obtained in a multipath acquisition stage, selecting the path with earliest arrival or strongest power as initial estimation of a first path, outputting the initial position of the first path, and providing an accurate initial value for subsequent fine synchronization.
  6. 6. The method for reconstructing a physical environment with integrated sense as set forth in claim 4, wherein in step S303, the frequency domain channel estimate is converted into a time domain power delay spectrum by IFFT according to the initial estimate of the first path; and correcting the head path position by adopting a differential channel power balance time delay estimation algorithm to obtain accurate arrival time.
  7. 7. The general sense integrated physical environment reconstruction method according to claim 1, wherein the parameterized digital twin model automatically extracts the position, size, posture and material properties of a scatterer through semantic segmentation, instance segmentation and three-dimensional reconstruction algorithm, and realizes millisecond-level environment capturing and automatic digital twin reconstruction of a complex dynamic scene.
  8. 8. The method for reconstructing the general sense integrated physical environment according to claim 1, wherein the optimized electromagnetic propagation calculation method combines a parallel ray tracing algorithm and an FDTD acceleration engine, and maintains an environment-channel association database at the same time, so as to support rapid mapping of an empirical model and a full-wave simulation result to improve calculation efficiency.
  9. 9. The method for reconstructing a physical environment with integrated sense of general according to claim 1, further comprising the steps of data processing and visualization, wherein the method comprises the steps of generating and dynamically updating indexes including channel impulse response, power delay spectrum, angle power spectrum and Doppler power spectrum, and visually presenting electromagnetic propagation panorama through a high-performance rendering engine.
  10. 10. The method for reconstructing a physical environment with integrated sense of general according to claim 1, wherein the multi-modal sensing means comprises unmanned aerial vehicle (unmanned aerial vehicle), camera vision, millimeter wave radar and traditional electromagnetic measurement.

Description

General sense integrated physical environment reconstruction method Technical Field The invention belongs to the technical field of wireless communication and positioning, and particularly relates to a general sense integrated physical environment reconstruction method. Background Wireless communication and positioning technologies mainly include conventional TOA based positioning and sensing technologies based on arrival time and positioning and sensing technologies based on classical channel models. The positioning and sensing technology based on the traditional TOA is limited by multipath interference and non-line-of-sight propagation, the indoor and outdoor ranging precision is generally low, the application requirements of high-precision complex scenes such as automatic driving, industrial robots and the like are difficult to meet, the classical channel model only depends on statistical parameters, the explicit modeling of the real scatterer geometry, materials and motion states is lacking, and the prediction error is high in the high-dynamic scene, so that the beam mismatch, the link interruption and the resource waste are caused. The disadvantages of the prior art are collated as follows: (1) The ranging accuracy is low, the ranging accuracy is limited by multipath interference and non-line-of-sight propagation, and the first-pass signal is often submerged by a strong reflection path in a complex environment, so that the delay estimation deviation reaches a plurality of sampling periods, the ranging accuracy is low, and the requirements of high-dynamic and high-safety scenes are difficult to meet. (2) And the channel prediction error is large, namely, the explicit modeling on the geometry of a scatterer, the dielectric characteristics of materials and the real-time motion trail is lacked, so that the beam pre-pointing failure, frequent interruption of switching and inefficiency of resource scheduling are caused. (3) The large-scale deployment is difficult, the special sensing signals or the additional hardware support are needed to be relied on, the deployment cost is high, the coverage is limited, and the large-scale application in the existing cellular network is difficult. Disclosure of Invention In order to solve the problems in the prior art, the invention provides a general sense integrated physical environment reconstruction method. The technical scheme adopted for solving the technical problems is that the general sense integrated physical environment reconstruction method takes electromagnetic wave propagation characteristics as an information carrier and comprises the following steps: s1, environment sensing, namely capturing multidimensional information of environment physical parameters, obstacles and mobile scatterers in real time through a multi-mode sensing means; S2, constructing a virtual physical environment, extracting key environment characteristics through a multi-source information fusion algorithm, and constructing a parameterized digital twin model, so that the shape, the position, the materials and the motion state of each object in the virtual physical environment are mapped with the real world in a high precision; S3, predicting electromagnetic wave propagation, calculating, evaluating and adjusting the electromagnetic wave propagation characteristics, adopting an optimized ultra-fast electric field calculation method, combining an environment-channel association database, and deducing electromagnetic propagation behaviors in real time in a reconstruction scene to output path parameters, wherein the path parameters comprise but are not limited to propagation directions, amplitudes, delays and phases; s4, forming a closed-loop autonomous system of perception-prediction-behavior, converting a prediction result into an optimization strategy, and reversely adjusting perception and prediction parameters to realize global and local two-stage closed-loop feedback. The physical environment reconstruction method is preferably realized based on a wireless environment prediction and reconstruction system, wherein the system adopts a five-layer architecture, namely an infrastructure layer, a resource layer, a network function layer, an application layer and a prediction plane crossing all layers, and a prediction plane embeds an environment sensing module, an electromagnetic wave propagation prediction module and a behavior module to form two-stage closed-loop feedback. Preferably, the first-stage closed-loop feedback penetrates through the whole physical environment layer, wherein an original perception sample is provided by a basic facility and a resource layer, a perception module extracts knowledge and reconstructs a virtual environment, a prediction module calculates and evaluates electromagnetic propagation characteristics, a behavior module generates an optimization strategy and feeds the optimization strategy back to a network function layer and an application layer, and global