CN110933685A - High-speed rail network coverage prediction method and device based on machine learning and ray tracing - Google Patents
High-speed rail network coverage prediction method and device based on machine learning and ray tracing Download PDFInfo
- Publication number
- CN110933685A CN110933685A CN202010072957.2A CN202010072957A CN110933685A CN 110933685 A CN110933685 A CN 110933685A CN 202010072957 A CN202010072957 A CN 202010072957A CN 110933685 A CN110933685 A CN 110933685A
- Authority
- CN
- China
- Prior art keywords
- speed rail
- machine learning
- scene
- ray tracing
- predicted value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W16/00—Network planning, e.g. coverage or traffic planning tools; Network deployment, e.g. resource partitioning or cells structures
- H04W16/22—Traffic simulation tools or models
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W16/00—Network planning, e.g. coverage or traffic planning tools; Network deployment, e.g. resource partitioning or cells structures
- H04W16/18—Network planning tools
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/06—Testing, supervising or monitoring using simulated traffic
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/08—Testing, supervising or monitoring using real traffic
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Artificial Intelligence (AREA)
- Monitoring And Testing Of Transmission In General (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the invention relates to a high-speed rail network coverage prediction method and a high-speed rail network coverage prediction device based on machine learning and ray tracing, wherein the method comprises the following steps: acquiring a three-dimensional electronic map of a target high-speed rail scene; based on a three-dimensional electronic map of a target high-speed rail scene, calculating a preliminary predicted value of each position measuring point in the target high-speed rail scene by using ray tracing simulation; correcting the preliminary predicted value through machine learning based on the actual measured value of each position measuring point under the same target high-speed rail scene and in combination with the preliminary predicted value of each position measuring point to obtain a correction factor of the preliminary predicted value; and predicting the receiving field intensity of the high-speed rail scene by using ray tracing simulation according to the correction factor of the preliminary predicted value. In the embodiment of the invention, a ray tracing simulation technology and depth reinforcement machine learning are utilized to provide more accurate input basis for scene correction, the application and deployment range is more universal, and the robustness is higher.
Description
Technical Field
The embodiment of the invention relates to the technical field of wireless, in particular to a high-speed rail network coverage prediction method and device based on machine learning and ray tracing.
Background
With the rapid development of wireless technology, the complexity of radio wave propagation presents a great challenge to the deployment and optimization of base stations in high-speed GSM-R wireless networks. Therefore, the method effectively predicts the coverage of the base station, is the basis for realizing the deployment planning and optimization of the wireless network base station, is also the premise for accurately positioning the problems existing in the existing network, can improve the site selection efficiency of the base station, ensure the communication and reduce the cost of deployment trial and error, and is the key point and the difficulty point of the planning and optimization of the wireless network.
The existing network optimization depends on repeated testing and manual debugging of a low-speed vehicle, so that the problems of low efficiency, long time, high cost, difficulty in popularization and the like exist, machine learning and ray tracing simulation technologies are urgently needed to replace testing vehicles and manual debugging, accurate GSM-R field intensity coverage prediction is provided for realizing accurate, efficient and one-key completion of high-speed rail intelligent network optimization, and urgent requirements and important application values are provided.
Disclosure of Invention
At least one embodiment of the invention provides a high-speed rail network coverage prediction method and device based on machine learning and ray tracing, and the problem of accuracy of high-speed rail GSM-R field coverage prediction is solved.
In a first aspect, an embodiment of the present invention provides a high-speed rail network coverage prediction method based on machine learning and ray tracing, where the method includes:
acquiring a three-dimensional electronic map of a target high-speed rail scene;
based on a three-dimensional electronic map of a target high-speed rail scene, calculating a preliminary predicted value of each position measuring point in the target high-speed rail scene by using ray tracing simulation;
correcting the preliminary predicted value through machine learning based on the actual measured value of each position measuring point under the same target high-speed rail scene and in combination with the preliminary predicted value of each position measuring point to obtain a correction factor of the preliminary predicted value;
and predicting the receiving field intensity of the high-speed rail scene by using ray tracing simulation according to the correction factor of the preliminary predicted value.
In some embodiments, the three-dimensional electronic map of the target high-speed rail scene comprises: three-dimensional geometric information and topographic information of the structure body, and a grid map and category identification of objects in the scene.
In some embodiments, calculating a preliminary prediction value for each position measurement point in a target high iron scene using ray tracing simulation based on a three-dimensional electronic map of the target high iron scene comprises:
and simulating the target high-speed rail scene and the arrangement by ray tracking according to the measuring route, the information of the associated base station and the position and the arrangement of the transceiver required by ray tracking, and extracting a preliminary predicted value.
In some embodiments, the preliminary prediction values include a complex electric field, propagation path, propagation mechanism, order, and structure associated with the path for each path.
In some embodiments, the above method further comprises:
and acquiring the actual measurement value of each measurement point under the high-speed rail scene with the same target by a drive test method.
In some embodiments, the actual measurement values include three-dimensional coordinate information of each measurement point in the same target high-speed rail scene and receiving field intensity information of a plurality of base stations detectable by each measurement point.
In some embodiments, the step of correcting the preliminary predicted value by machine learning based on the actual measured value of each position measurement point in the high-speed rail scene with the same target in combination with the preliminary predicted value of each position measurement point to obtain a correction factor of the preliminary predicted value includes:
obtaining a field intensity error based on the actual measured field intensity and the preliminary predicted field intensity;
determining a current prediction error data characteristic set according to the field intensity error and a first threshold;
and based on the current prediction error data feature set and the historical prediction error data feature set, utilizing deep reinforcement machine learning to realize singular value judgment.
In some embodiments, the singular value decision is implemented by using deep enhanced machine learning based on the current prediction error data feature set and the historical prediction error data feature set, and includes:
judging whether a singular value exists or not or whether a singular data set has change or not, if not, finishing correction; and if not, restarting correction, and adding the current prediction error data characteristic set into the historical prediction error data characteristic set.
In a second aspect, an embodiment of the present invention further provides a high-speed rail network coverage prediction apparatus based on machine learning and ray tracing, including: the device comprises an electronic map acquisition module, a multipath information acquisition module, a correction module and a field intensity prediction module;
the electronic map acquisition module is used for acquiring a three-dimensional electronic map of a target high-speed rail scene;
the multi-path information acquisition module is used for calculating a preliminary predicted value of each position measuring point in the target high-speed rail scene by using ray tracing simulation based on a three-dimensional electronic map of the target high-speed rail scene;
the correction module is used for correcting the preliminary predicted value through machine learning based on the actual measured value of each position measuring point under the high-speed rail scene with the same target and the preliminary predicted value of each position measuring point to obtain a correction factor of the preliminary predicted value;
and the field intensity prediction module is used for predicting the receiving field intensity of the high-speed rail scene by using ray tracing simulation according to the correction factor of the preliminary prediction value.
The high-speed rail network coverage prediction method provided by the embodiment of the invention comprises the steps of obtaining a three-dimensional electronic map of a target high-speed rail scene, calculating a preliminary predicted value of each measuring point in the target high-speed rail scene by using ray tracing simulation according to the three-dimensional electronic map of the target high-speed rail scene, recording an actual measured value of each measuring point in the same target high-speed rail scene, correcting the preliminary predicted value by combining the preliminary predicted value of each measuring point through machine learning to obtain a correction factor of the preliminary predicted value, and finally performing high-speed rail scene receiving field intensity prediction by using ray tracing simulation according to the correction factor of the preliminary predicted value, so that the correction of the high-speed rail scene is more accurate, the application and deployment range is more universal, and the robustness is higher.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art according to the drawings.
Fig. 1 is a flowchart of a high-speed rail network coverage prediction method based on machine learning and ray tracing according to an embodiment of the present invention;
fig. 2 is a diagram of a macro station route measurement scenario and a route provided in an embodiment of the present invention;
FIG. 3 is a flowchart illustrating a calibration process according to an embodiment of the present invention;
FIG. 4 is a diagram of a structural material identifier corresponding to a changed singular value according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of a high-speed rail network coverage prediction apparatus based on machine learning and ray tracing according to an embodiment of the present invention.
Detailed Description
In order that the above objects, features and advantages of the present invention can be more clearly understood, the present invention will be further described in detail with reference to the accompanying drawings and examples. It is to be understood that the embodiments described are only a few embodiments of the present invention, and not all embodiments. The specific embodiments described herein are merely illustrative of the invention and are not to be construed as limiting the invention. All other embodiments, which can be derived by a person skilled in the art from the described embodiments of the invention, are within the scope of the invention.
It is noted that, in this document, relational terms such as "first" and "second," and the like, may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions.
Fig. 1 is a flowchart of a high-speed rail network coverage prediction method according to an embodiment of the present invention, as shown in fig. 1, the method includes the following steps:
step 101: and acquiring a three-dimensional electronic map of the target high-speed rail scene.
Specifically, in this step, a three-dimensional electronic map of a target high-speed rail scene is obtained by ray tracing, and the electronic map generally includes three-dimensional geometric information and topographic information of a structure, a grid map of objects in the scene, and a category identifier, for example: 1. tall building, 2. short building, 3. residential building and 4. office building.
Taking the GSM-R network coverage of a high-speed rail as an example, the three-dimensional electronic map for obtaining a target high-speed rail scene is a three-dimensional electronic map for obtaining a surrounding scene in the running process of the high-speed rail, and the structure body of the three-dimensional electronic map can include surrounding residential buildings, office buildings, vehicles coming and going, and the like, and grid maps and category identifiers of all objects in the surrounding scene.
Step 102: and calculating the preliminary predicted value of each position measuring point in the target high-speed rail scene by using ray tracing simulation based on the three-dimensional electronic map of the target high-speed rail scene.
As described in step 102, a preliminary prediction value for each position measurement point is calculated using ray tracing simulations, including at least complex electric fields, propagation paths, propagation mechanisms, orders, structures associated with the paths, and the like.
Step 103: and correcting the preliminary predicted value through machine learning based on the actual measured value of each position measuring point under the high-speed rail scene with the same target and the preliminary predicted value of each position measuring point to obtain a correction factor of the preliminary predicted value.
Specifically, in this step, combine actual measurement value and preliminary predicted value, rectify preliminary predicted value through machine learning, obtain the correction factor of preliminary predicted value, in its correction process, obtain the material property of the structure body of laminating reality, wherein the material property includes material sign, propagation model coefficient. Wherein the propagation model coefficients include: equivalent thickness, roughness, dielectric constant, transmission coefficient, diffraction coefficient, scattering gain, etc., as described below;
and obtaining a propagation model fitting the reality, wherein the propagation model refers to: reflection, scattering, diffraction, transmission, and the like. The propagation model corresponding to reflection is a Fresnel model and a kirchhoff model, the propagation model corresponding to scattering is a directional scattering model, the propagation model corresponding to diffraction is Deygout, and the propagation model corresponding to transmission is penetration loss.
The corresponding relation between the propagation model and the propagation parameters is as follows: fresnel formula-dielectric constant, directional scattering model-roughness, scattering gain, diffraction-diffraction coefficient, transmission-equivalent thickness, transmission coefficient, dielectric constant, as described below;
the method comprises the steps of obtaining actual measurement values of partial areas of a target high-speed rail scene based on means such as drive tests, wherein the actual measurement values comprise three-dimensional coordinate information of measurement positions and receiving field intensity information of a plurality of base stations which can be detected by each measurement position.
Step 104: and predicting the receiving field intensity of the high-speed rail scene by using ray tracing simulation according to the correction factor of the preliminary predicted value.
In some embodiments, a preliminary prediction value for each measurement point in a target high iron scene is calculated using ray tracing simulation based on a three-dimensional electronic map of the target high iron scene, the method comprising:
simulating a target high-speed rail scene and deployment through ray tracking according to an actual measurement route, associated base station information and the position and deployment of a transceiver required by ray tracking, and extracting a preliminary predicted value;
the preliminary predicted value comprises a plurality of electric fields, propagation paths, propagation mechanisms, orders and structural bodies related to the paths.
Fig. 2 is a diagram of a macro station route measurement scenario and a route according to an embodiment of the present invention, and as shown in fig. 2, the macro station route measurement scenario and an actual measurement route are shown in detail.
In fig. 2, 1 is the position of the macro station, and the drive test equipment randomly measures the actual measurement value anywhere within a certain range, and the random route is shown as 2.
Optionally, in some embodiments, the actual measurement value of each measurement point in the same target high-speed rail scene is obtained through a drive test means.
In some embodiments, based on the actual measurement value of each measurement point in the high-iron scene with the same target, the preliminary prediction value of each measurement point is combined, and the preliminary prediction value is corrected through machine learning, so as to obtain a correction factor of the preliminary prediction value:
obtaining a field intensity error based on the actual measured field intensity and the preliminary predicted field intensity;
determining a current prediction error data characteristic set according to the field intensity error and a first threshold;
and based on the current prediction error data feature set and the historical prediction error data feature set, utilizing deep reinforcement machine learning to realize singular value judgment.
Optionally, in some embodiments, the singular value decision is implemented, including:
judging whether a singular value exists or not or whether a singular data set has change or not, if not, finishing correction; otherwise, restarting correction, and adding the current prediction error data characteristic set into the historical prediction error data characteristic set;
wherein, the historical prediction error data characteristic set does not exist at the beginning of the correction process.
Fig. 3 is a flowchart of a calibration method according to an embodiment of the present invention, as shown in fig. 3, including the following steps:
step 201: and calculating the error of the preliminary predicted value and the actual measured value.
Defining the position and deployment of a transceiver required by ray tracking according to the measurement route and the information of the associated base station, simulating the current high-speed rail scene and deployment by using the ray tracking, extracting a preliminary predicted value, and calculating the error Er between simulation and actual drive test field intensity;
fitting propagation parameters for a propagation model and a structural body corresponding to each multipath by using an unsupervised gradient descent method based on a reflection, scattering, transmission and diffraction multipath propagation model by taking minimum Er as a target;
finally, the optimal propagation parameters of the related structural body materials are obtained under the constraint condition of the current scene geometric model through the steps, so that the optimal propagation parameters are local optimal solutions; the multipath propagation model is as follows.
The geometric tracking of the direct multipath type is free space propagation, an electromagnetic calculation model is Friis equalization, and no correctable electromagnetic parameters exist; the geometric tracking of the reflection multipath type is Snell reflection law, the electromagnetic calculation model is a Fresnel formula, and the correctable electromagnetic parameter is dielectric constant; the geometric tracking of the transmission multipath type is Snell transmission law, the electromagnetic calculation model is Fresnel formula and penetration loss, and the correctable electromagnetic parameters are equivalent thickness, transmission coefficient and dielectric constant; the geometric tracking of the scattering multipath type is a binning splitting method, the electromagnetic calculation model is a directional scattering model, and electromagnetic parameters can be corrected into roughness and scattering gain; the geometric tracking of the type of multipath diffracted is a Deygout model, the electromagnetic calculation model is a Deygout model, and the correctable electromagnetic parameters are the coefficients of diffraction.
Step 202: and (3) singular value judgment: setting a first threshold (PT) of an error Er between the calculation simulation field intensity and the actual drive test field intensity, extracting a data set S of which the Er is larger than the PT in the step 201, wherein the data set S comprises the Er, a three-dimensional coordinate corresponding to the position of a transceiver, and a primary predicted value obtained in the step 103, wherein the primary predicted value comprises multipath types, paths, passing structural bodies and the like, combining a historical prediction error data feature set and a current prediction feature set, and realizing singular value judgment by using a deep reinforcement learning method, wherein the specific process is as follows:
(1) selecting learning characteristics from dimensions such as the length of a continuous interval of a historical prediction error data characteristic set and a current data set S, the orientation relation of a transceiver, error distribution, the composition of a multipath mechanism, the composition of a structural body and a material and the like;
(2) and obtaining the judgment whether the singular value exists or not according to the magnitude relation between Er and PT of the current data set S. Combining the data set characteristics with the historical prediction error larger than PT and the characteristics of the current data set S to obtain the current state StWhether to proceed with "corrective" action atProbability p (a) oft|st) And after taking the action, taking the action atAnd the current state stReturn to the next state st+1Probability p(s) oft+1|at,st)；
(3) From (2), an action state sequence A = { (a)t,st) T =1,2,3, …, T }, and an arbitrary state s can be obtainedtTake action atProbability of (p), (t), corresponding current state and action (a)t,st) The benefits r (t) below, and take a series of actions { (a)t,st) Total benefit r (t) after t =1,2,3, …, t }.
Wherein R (m) represents a state smTake action amCorresponding benefit, p (m) denotes the state smTake action amCorresponding probability, if action a is takenmThen, the error Er between the simulation result and the measurement result is reduced, and the corresponding benefit r (m) of the action is obtained>0, i.e. reward; on the contrary, if Er is increased, the corresponding benefit r (m)<0, i.e. penalty.
(4) A singular value decision second threshold Et is set, and the benefit r (t) is compared with the decision second threshold Et. If | r (t) | > Et, the scene parameter correction is performed (step 203). If | r (t) | < Et, the correction is terminated, and the correction factor extracted by the deep reinforcement learning is stored.
The method described in step 202 may reduce the sensitivity to the influence of dynamic structures in the measurement system or measurement scene, such as vehicles, people, etc. passing randomly, so as to accurately locate the static structures with singular influence caused in the three-dimensional electronic map.
And step 203, changing the material type of the structural body related to the singular value.
If the singular value exists in the step 202, the material identifier of the structural body corresponding to the singular value is created and modified in the step, so that the material identifier is different from the original identifier and the structural body with the original identifier.
Step 204: and finishing the correction and saving the correction factor at the position.
As shown in fig. 4, the modified structure is considered as a new material, the new scene model replaces the old model, the next round of correction is performed, step 201 is executed again, the feature set of the current predicted singular value is added into the historical singular set, and the current singular value decision strategy is added into the historical strategy, so as to perfect the reward and punishment of machine learning.
Fig. 5 is a schematic structural diagram of a high-speed rail network coverage prediction apparatus based on machine learning and ray tracing according to an embodiment of the present invention, as shown in fig. 5, the apparatus includes: an electronic map acquisition module 301, a multipath information acquisition module 302, a correction module 303 and a field intensity prediction module 304;
the electronic map acquisition module is used for acquiring a three-dimensional electronic map of a target high-speed rail scene;
the multi-path information acquisition module is used for calculating a preliminary predicted value of each position measuring point in the target high-speed rail scene by using ray tracing simulation based on a three-dimensional electronic map of the target high-speed rail scene;
the correction module is used for correcting the preliminary predicted value through machine learning based on the actual measured value of each position measuring point under the high-speed rail scene with the same target and the preliminary predicted value of each position measuring point to obtain a correction factor of the preliminary predicted value;
and the field intensity prediction module is used for predicting the receiving field intensity of the high-speed rail scene by using ray tracing simulation according to the correction factor of the preliminary prediction value.
In the embodiment of the invention, a ray tracing simulation technology and depth reinforcement machine learning are utilized to provide more accurate input basis for scene correction, the application and deployment range is more universal, and the robustness is higher.
The apparatus disclosed in the above embodiments can implement the processes of the methods disclosed in the above method embodiments, and in order to avoid repetition, the details are not described here again.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
Those skilled in the art will appreciate that although some embodiments described herein include some features included in other embodiments instead of others, combinations of features of different embodiments are meant to be within the scope of the invention and form different embodiments.
Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations fall within the scope defined by the appended claims.
Claims (9)
1. A high-speed rail network coverage prediction method based on machine learning and ray tracing is characterized by comprising the following steps:
acquiring a three-dimensional electronic map of a target high-speed rail scene;
based on the three-dimensional electronic map of the target high-speed rail scene, calculating a preliminary predicted value of each position measuring point in the target high-speed rail scene by using ray tracing simulation;
correcting the preliminary predicted value through machine learning based on the actual measured value of each position measuring point under the same target high-speed rail scene and in combination with the preliminary predicted value of each position measuring point to obtain a correction factor of the preliminary predicted value;
and predicting the receiving field intensity of the high-speed rail scene by using ray tracing simulation according to the correction factor of the preliminary predicted value.
2. The machine learning and ray tracing based high-speed rail network coverage prediction method according to claim 1, wherein the three-dimensional electronic map of the target high-speed rail scene comprises: three-dimensional geometric information and topographic information of the structure body, and a grid map and category identification of objects in the scene.
3. The machine learning and ray tracing based high-speed rail network coverage prediction method according to claim 1, wherein the calculating a preliminary prediction value of each position measurement point in the target high-speed rail scene using ray tracing simulation based on the three-dimensional electronic map of the target high-speed rail scene comprises:
and simulating the target high-speed rail scene and the arrangement by ray tracking according to the measuring route, the information of the associated base station and the position and the arrangement of the transceiver required by ray tracking, and extracting the preliminary predicted value.
4. The machine learning and ray tracing based high-speed rail network coverage prediction method according to claim 3, wherein the preliminary prediction values include complex electric fields, propagation paths, propagation mechanisms, orders and structures associated with the paths.
5. The high-speed rail network coverage prediction method based on machine learning and ray tracing as claimed in claim 1, characterized in that the actual measurement value of each position measurement point under the same target high-speed rail scene is obtained by a drive test means.
6. The method according to claim 5, wherein the actual measurement values include three-dimensional coordinate information of each position measurement point in the same target high-speed rail scene and the received field strength information of a plurality of base stations detectable by each position measurement point.
7. The method for predicting the coverage of the high-speed rail network based on machine learning and ray tracing according to claim 1, wherein the step of correcting the preliminary predicted value by machine learning based on the actual measured value of each position measuring point in the same target high-speed rail scene in combination with the preliminary predicted value of each position measuring point to obtain a correction factor of the preliminary predicted value comprises:
obtaining a field intensity error based on the actual measured field intensity and the preliminary predicted field intensity;
determining a current prediction error data characteristic set according to the field intensity error and a first threshold;
and based on the current prediction error data feature set and the historical prediction error data feature set, utilizing deep reinforcement machine learning to realize singular value judgment.
8. The machine learning and ray tracing-based high-speed rail network coverage prediction method according to claim 7, wherein the performing singular value decision by using deep enhanced machine learning based on the current prediction error data feature set and the historical prediction error data feature set comprises:
judging whether a singular value exists or not or whether a singular data set has change or not, if not, finishing the correction; otherwise, the correction is restarted, and the current prediction error data characteristic set is added into the historical prediction error data characteristic set.
9. A high-speed rail network coverage prediction device based on machine learning and ray tracing, comprising:
the electronic map acquisition module is used for acquiring a three-dimensional electronic map of a target high-speed rail scene;
the multipath information acquisition module is used for calculating a preliminary predicted value of each position measuring point in the target high-speed rail scene by using ray tracing simulation based on the three-dimensional electronic map of the target high-speed rail scene;
the correction module is used for correcting the preliminary predicted value through machine learning based on the actual measured value of each position measuring point under the high-speed rail scene with the same target and in combination with the preliminary predicted value of each position measuring point to obtain a correction factor of the preliminary predicted value;
and the field intensity prediction module is used for predicting the receiving field intensity of the high-speed rail scene by using ray tracing simulation according to the correction factor of the preliminary prediction value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010072957.2A CN110933685B (en) | 2020-01-22 | 2020-01-22 | High-speed rail network coverage prediction method and device based on machine learning and ray tracing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010072957.2A CN110933685B (en) | 2020-01-22 | 2020-01-22 | High-speed rail network coverage prediction method and device based on machine learning and ray tracing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110933685A true CN110933685A (en) | 2020-03-27 |
| CN110933685B CN110933685B (en) | 2020-06-05 |
Family
ID=69854421
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010072957.2A Active CN110933685B (en) | 2020-01-22 | 2020-01-22 | High-speed rail network coverage prediction method and device based on machine learning and ray tracing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110933685B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111162847A (en) * | 2020-04-02 | 2020-05-15 | 北京中铁建电气化设计研究院有限公司 | Alignment method and device for high-speed rail network directional antenna and electronic equipment |
| CN113938895A (en) * | 2021-09-16 | 2022-01-14 | 中铁第四勘察设计院集团有限公司 | Method and device for predicting railway wireless signal, electronic equipment and storage medium |
| CN114297747A (en) * | 2021-12-06 | 2022-04-08 | 上海中铁通信信号测试有限公司 | Subway tunnel mixed channel modeling method and electronic terminal |
| CN114422061A (en) * | 2022-03-31 | 2022-04-29 | 中铁第四勘察设计院集团有限公司 | Adaptive prediction method for wireless signal propagation in railway environment |
| CN114448474A (en) * | 2020-11-05 | 2022-05-06 | 中国移动通信集团设计院有限公司 | Multi-antenna channel matrix prediction method and device and electronic equipment |
| CN115065955A (en) * | 2022-08-18 | 2022-09-16 | 中国铁建电气化局集团有限公司 | High-speed rail 5G wireless communication network coverage planning method, device, equipment and medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103365962A (en) * | 2013-06-19 | 2013-10-23 | 山东润谱通信工程有限公司 | Building and calibrating method for construction material wireless propagation loss parameter database |
| CN103702338A (en) * | 2013-12-24 | 2014-04-02 | 山东润谱通信工程有限公司 | Method for rapidly establishing indoor wireless signal fingerprint database |
| CN104363616A (en) * | 2014-10-27 | 2015-02-18 | 山东润谱通信工程有限公司 | Method for predicting indoor three-dimensional space field strength from outdoor to indoor with propagation models |
| CN104660349A (en) * | 2014-10-27 | 2015-05-27 | 山东润谱通信工程有限公司 | Method for predicting outdoor three-dimensional space field intensity through expanded COST-231-Walfisch-Ikegami propagation model |
| CN109217955A (en) * | 2018-07-13 | 2019-01-15 | 北京交通大学 | Wireless environment electromagnetic parameter approximating method based on machine learning |
| CN110602736A (en) * | 2019-07-10 | 2019-12-20 | 北京交通大学 | Method and device for field intensity prediction and computer equipment |
-
2020
- 2020-01-22 CN CN202010072957.2A patent/CN110933685B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103365962A (en) * | 2013-06-19 | 2013-10-23 | 山东润谱通信工程有限公司 | Building and calibrating method for construction material wireless propagation loss parameter database |
| CN103702338A (en) * | 2013-12-24 | 2014-04-02 | 山东润谱通信工程有限公司 | Method for rapidly establishing indoor wireless signal fingerprint database |
| CN104363616A (en) * | 2014-10-27 | 2015-02-18 | 山东润谱通信工程有限公司 | Method for predicting indoor three-dimensional space field strength from outdoor to indoor with propagation models |
| CN104660349A (en) * | 2014-10-27 | 2015-05-27 | 山东润谱通信工程有限公司 | Method for predicting outdoor three-dimensional space field intensity through expanded COST-231-Walfisch-Ikegami propagation model |
| CN109217955A (en) * | 2018-07-13 | 2019-01-15 | 北京交通大学 | Wireless environment electromagnetic parameter approximating method based on machine learning |
| CN110602736A (en) * | 2019-07-10 | 2019-12-20 | 北京交通大学 | Method and device for field intensity prediction and computer equipment |
Non-Patent Citations (2)
| Title |
|---|
| TAO ZHOU等: "Channel Modeling for Future High-Speed Railway", 《IEEE ACCESS》 * |
| 官科: "轨道交通场景电波传播建模理论与方法研究", 《万方学位论文库》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111162847A (en) * | 2020-04-02 | 2020-05-15 | 北京中铁建电气化设计研究院有限公司 | Alignment method and device for high-speed rail network directional antenna and electronic equipment |
| CN114448474A (en) * | 2020-11-05 | 2022-05-06 | 中国移动通信集团设计院有限公司 | Multi-antenna channel matrix prediction method and device and electronic equipment |
| CN114448474B (en) * | 2020-11-05 | 2022-11-22 | 中国移动通信集团设计院有限公司 | Multi-antenna channel matrix prediction method and device and electronic equipment |
| CN113938895A (en) * | 2021-09-16 | 2022-01-14 | 中铁第四勘察设计院集团有限公司 | Method and device for predicting railway wireless signal, electronic equipment and storage medium |
| CN113938895B (en) * | 2021-09-16 | 2023-09-05 | 中铁第四勘察设计院集团有限公司 | Prediction method and device for railway wireless signal, electronic equipment and storage medium |
| CN114297747A (en) * | 2021-12-06 | 2022-04-08 | 上海中铁通信信号测试有限公司 | Subway tunnel mixed channel modeling method and electronic terminal |
| CN114422061A (en) * | 2022-03-31 | 2022-04-29 | 中铁第四勘察设计院集团有限公司 | Adaptive prediction method for wireless signal propagation in railway environment |
| CN114422061B (en) * | 2022-03-31 | 2022-07-19 | 中铁第四勘察设计院集团有限公司 | Adaptive prediction method for wireless signal propagation in railway environment |
| CN115065955A (en) * | 2022-08-18 | 2022-09-16 | 中国铁建电气化局集团有限公司 | High-speed rail 5G wireless communication network coverage planning method, device, equipment and medium |
| CN115065955B (en) * | 2022-08-18 | 2022-12-13 | 中国铁建电气化局集团有限公司 | High-speed rail 5G wireless communication network coverage planning method, device, equipment and medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110933685B (en) | 2020-06-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110933685B (en) | High-speed rail network coverage prediction method and device based on machine learning and ray tracing | |
| CN103841640B (en) | NLOS base station identifying and positioning method based on positioning position residual error | |
| Li et al. | A feature-scaling-based $ k $-nearest neighbor algorithm for indoor positioning systems | |
| CN103379441B (en) | A kind of indoor orientation method based on curve and location finding | |
| CN102984745B (en) | Combined estimation method for Wi-Fi AP (wireless fidelity access point) position and path loss model | |
| CN103997783B (en) | A kind of outdoor cluster match localization method and device | |
| CN111416676B (en) | High-speed rail railway crossing and merging section field strength prediction method based on ray tracking | |
| CN104703143A (en) | Indoor positioning method based on WIFI signal strength | |
| CN106792465A (en) | A kind of indoor fingerprint map constructing method based on mass-rent fingerprint | |
| CN106507411A (en) | A kind of LTE works ginseng automatic inspection method based on MR | |
| CN102395195B (en) | Method for raising indoor positioning precision under non-line-of-sight environment | |
| CN101180550A (en) | Enhanced mobile location method and system | |
| CN112469066B (en) | 5G network coverage evaluation method and device | |
| CN101909241B (en) | System and method for identifying the position of mobile terminals | |
| US20100137005A1 (en) | Method for positioning portable communication device | |
| Pinto et al. | Robust RSSI-based indoor positioning system using K-means clustering and Bayesian estimation | |
| CN104661232B (en) | The AP method for arranging limited substantially based on Fisher's information matrix fingerprint positioning precision | |
| CN104581945A (en) | WLAN indoor positioning method for distance constraint based semi-supervised APC clustering algorithm | |
| CN101541079B (en) | Traveling carriage positioning method | |
| CN103561412A (en) | Channel associated shadow fading construction method based on stationary random process | |
| Ji et al. | Accurate Long‐Term Evolution/Wi‐Fi hybrid positioning technology for emergency rescue | |
| CN101373216A (en) | Method for positioning portable communication device | |
| CN112995893A (en) | Fingerprint positioning method, system, server and storage medium | |
| CN102469477B (en) | Network optimization method, apparatus thereof, and system thereof | |
| CN105828342A (en) | Method and device for confirming neighboring relation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |