CN-121980937-A - Three-dimensional wind field correction and greenhouse gas tracing method

CN121980937ACN 121980937 ACN121980937 ACN 121980937ACN-121980937-A

Abstract

The invention discloses a three-dimensional wind field correction and greenhouse gas tracing method, and relates to the technical field of vehicle-mounted environment monitoring and gas tracing. The three-dimensional wind field correction and greenhouse gas tracing method comprises the steps of constructing a vehicle-mounted multi-source sensing fusion acquisition system, synchronously acquiring three-dimensional wind original data, carrier motion state data and greenhouse gas concentration data, acquiring geographic coordinate system three-dimensional wind field correction data through layered correction processing comprising time synchronization, coordinate transformation, carrier motion error correction and shadow effect compensation, and positioning and quantifying a greenhouse gas emission source based on the geographic coordinate system three-dimensional wind field correction data and the greenhouse gas concentration data through an optimized diffusion model and inversion algorithm.

Inventors

YUAN SONG
HE JUNFENG
WEI MIN
LI SHENG
LI MINGXING

Assignees

合肥综合性科学中心环境研究院

Dates

Publication Date: 20260505
Application Date: 20260123

Claims (10)

1. A three-dimensional wind field correction and greenhouse gas tracing method is characterized by comprising the following steps: Constructing a vehicle-mounted multi-source sensing fusion acquisition system, and synchronously acquiring three-dimensional wind original data, carrier motion state data and greenhouse gas concentration data; based on the carrier motion state data, carrying out layering correction processing comprising time synchronization, coordinate transformation, carrier motion error correction and shadow effect compensation on the three-dimensional wind original data to obtain geographic coordinate system three-dimensional wind field correction data; and positioning and quantitatively tracing a greenhouse gas emission source through an optimized diffusion model and an inversion algorithm based on the three-dimensional wind field correction data of the geographic coordinate system and the greenhouse gas concentration data.
2. The three-dimensional wind field correction and greenhouse gas tracing method according to claim 1, wherein the coordinate transformation is realized by constructing a transformation matrix from a carrier coordinate system to a geographic coordinate system, and the expression of the transformation matrix is: ; Wherein, the 、、 Respectively a rotation matrix around the Z axis, the X axis and the Y axis of the carrier coordinate system, 、、 Respectively the real-time carrier yaw angle, pitch angle and roll angle obtained from the carrier motion state data.
3. The three-dimensional wind field correction and greenhouse gas tracing method according to claim 2, wherein the carrier movement error correction comprises: and correcting translational errors, namely calculating a first intermediate wind speed vector after eliminating the influence of the translational speed of the carrier by the following formula: ; Wherein, the For the wind velocity vector in the three-dimensional wind raw data in the carrier coordinate system, Is a translational velocity vector in the carrier motion state data; And correcting rotation errors, namely calculating a second intermediate wind speed vector after further eliminating the influence of carrier rotation by the following formula: ; Wherein, the Is an angular velocity vector in the carrier motion state data, The detection distance vector of the wind measuring sensor is obtained.
4. A three-dimensional wind-field correction and greenhouse-gas tracing method according to claim 3, wherein said shadow effect compensation uses a trained random forest regression model for said second intermediate wind velocity vector And compensating to output final three-dimensional wind field correction data of the geographic coordinate system, wherein training data of the random forest regression model is generated based on fluid mechanics simulation and real vehicle test comparison data.
5. The three-dimensional wind field correction and greenhouse gas tracing method according to claim 1, wherein the optimized diffusion model is an improved gaussian diffusion model, and the function expression of the predicted concentration distribution is: ; Wherein, the Is a space point The predicted gas concentration at which to proceed, In order to discharge the source intensity, The horizontal average wind speed in the data is corrected for the three-dimensional wind field of the geographic coordinate system, 、 The horizontal and vertical diffusion parameters are respectively, Is the effective height of the emission source.
6. The method for three-dimensional wind field correction and greenhouse gas tracing as defined in claim 5, wherein in said modified Gaussian diffusion model, the input wind speed is corrected according to altitude change according to wind speed profile exponential law, and said diffusion parameters are 、 And dynamically adjusting the wind direction stability and the atmosphere stability in the three-dimensional wind field correction data of the geographic coordinate system.
7. The method according to claim 5, wherein the inversion algorithm is a particle swarm optimization algorithm that iteratively solves for the position coordinates, emission intensity, and emission time characteristic parameters of the emission source with a root mean square error between the predicted concentration of the modified gaussian diffusion model and the measured concentration in the greenhouse gas concentration data as an objective function.
8. The method for correcting a three-dimensional wind field and tracing greenhouse gases according to claim 1, wherein the time synchronization ensures the sampling start synchronization of each sensor through a hardware trigger pulse, and adopts a software timestamp interpolation method to align data streams, so that the time deviation among the three-dimensional wind raw data, carrier motion state data and greenhouse gas concentration data is less than 10 milliseconds.
9. The three-dimensional wind field correction and greenhouse gas tracing method according to claim 1, further comprising a tracing result verification step of introducing an interpretable machine learning model, quantifying contribution weights of the three-dimensional wind field correction data, traffic flow and building layout of the geographic coordinate system to concentration distribution, and fusing image identification information to perform cross verification and optimization on a preliminary tracing result.
10. The three-dimensional wind field correction and greenhouse gas tracing method according to claim 1, wherein in the vehicle-mounted multi-source sensing fusion acquisition system, a laser wind-finding radar and an inertial measurement unit are rigidly co-located at the center of a vehicle roof, and the main detection axis of the laser wind-finding radar is consistent with the longitudinal axis direction of the vehicle.

Description

Three-dimensional wind field correction and greenhouse gas tracing method Technical Field The invention relates to the technical field of vehicle-mounted environment monitoring and gas tracing, in particular to a three-dimensional wind field correction and greenhouse gas tracing method. Background The vehicle-mounted greenhouse gas navigation monitoring system has the advantages of high space-time resolution and capability of dynamically covering a large area, and becomes one of core technical means for monitoring the emission of greenhouse gases in cities and areas. The system can realize rapid positioning and tracing of the emission hot spot area by synchronously collecting multi-source data such as greenhouse gas concentration, meteorological parameters, geographic information and the like. The three-dimensional wind speed and direction serve as key driving parameters of greenhouse gas diffusion and propagation, and the measurement accuracy directly determines the inversion accuracy of the traceability model. In a vehicle-mounted mobile scene, three-dimensional wind speed and direction measurement faces a plurality of inherent problems: Firstly, the carrier movement interference is remarkable, and the translation (uniform speed/acceleration linear movement) and rotation (steering, pitching and rolling) of the vehicle can enable the measured value of the wind measuring sensor to be overlapped with the speed and posture offset of the carrier, so that the original wind field data can not directly reflect the real atmospheric movement state; secondly, the attitude of the sensor is dynamically changed, and the change of yaw angle and pitch angle of vehicle-mounted wind measuring equipment (such as a laser wind measuring radar and an ultrasonic anemometer) along with the running process of a vehicle can cause the deviation of a laser beam pointing or ultrasonic detection path, so that an angle measurement error is introduced; thirdly, the influence of sensor shadow effect and vehicle body turbulence exists, the structural design and vehicle-mounted mounting position of the wind measuring equipment are easy to generate airflow shielding, meanwhile, the turbulence formed by the high-speed movement of the vehicle body can change the local wind field distribution, and the measurement deviation is further increased; Fourth, the existing correction method has limitations, the traditional wind field correction algorithm based on the fixed platform cannot be adapted to the vehicle-mounted dynamic scene, only a single motion error (such as only correcting translational speed) is considered in part of the mobile platform correction scheme, the comprehensive influence of rotation, gesture and shadow effect is ignored, the problems of large calculation amount, poor real-time performance, no consideration of time delay of a multi-source sensor and the like exist, and the dual requirements of the navigation monitoring on the real-time performance and the precision are difficult to meet. The error of the wind speed and direction measurement can be directly transmitted to a greenhouse gas tracing link. The existing tracing method is mostly based on a Gaussian diffusion model, an Euler-Lagrange particle diffusion model and the like, and the core input parameter of the tracing method is wind speed and direction. If the wind field data have deviation, the simulation distortion of the diffusion path of the emission source and the inversion deviation of the emission intensity are too large, and the accurate positioning and quantitative analysis of the greenhouse gas emission source cannot be realized. Disclosure of Invention Aiming at the defects of the prior art, the invention provides a three-dimensional wind field correction and greenhouse gas tracing method, which solves the problems that the three-dimensional wind speed and direction measurement in the existing vehicle-mounted greenhouse gas navigation system is easily influenced by carrier movement, attitude change and sensor interference, so that the measurement precision is low and the greenhouse gas tracing is inaccurate. The three-dimensional wind field correction and greenhouse gas tracing method comprises the following steps of constructing a vehicle-mounted multi-source sensing fusion acquisition system, synchronously acquiring three-dimensional wind original data, carrier motion state data and greenhouse gas concentration data, carrying out layered correction processing on the three-dimensional wind original data, comprising time synchronization, coordinate transformation, carrier motion error correction and shadow effect compensation, based on the carrier motion state data, to obtain three-dimensional wind field correction data of a geographic coordinate system, and carrying out positioning and quantification tracing on a greenhouse gas emission source through an optimized diffusion model and an inversion algorithm based on the three-dimensional wind field correctio