CN112883643A - Simulation test system for ship carrying laser sensing module - Google Patents
Simulation test system for ship carrying laser sensing module Download PDFInfo
- Publication number
- CN112883643A CN112883643A CN202110176950.XA CN202110176950A CN112883643A CN 112883643 A CN112883643 A CN 112883643A CN 202110176950 A CN202110176950 A CN 202110176950A CN 112883643 A CN112883643 A CN 112883643A
- Authority
- CN
- China
- Prior art keywords
- module
- data
- class
- data point
- ship
- 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.)
- Pending
Links
- 238000012360 testing method Methods 0.000 title claims abstract description 53
- 238000004088 simulation Methods 0.000 title claims abstract description 41
- 238000012545 processing Methods 0.000 claims abstract description 25
- 230000004888 barrier function Effects 0.000 claims abstract description 7
- 239000013598 vector Substances 0.000 claims description 43
- 230000008447 perception Effects 0.000 claims description 19
- 230000006870 function Effects 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 9
- 238000005259 measurement Methods 0.000 claims description 9
- 238000000034 method Methods 0.000 claims description 7
- 238000001914 filtration Methods 0.000 claims description 5
- 230000008569 process Effects 0.000 claims description 4
- 238000004422 calculation algorithm Methods 0.000 claims description 3
- 239000000126 substance Substances 0.000 claims description 2
- 230000003287 optical effect Effects 0.000 claims 1
- 238000001514 detection method Methods 0.000 description 13
- 230000000694 effects Effects 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 230000011218 segmentation Effects 0.000 description 4
- 230000003068 static effect Effects 0.000 description 4
- 238000011161 development Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000010998 test method Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 230000036039 immunity Effects 0.000 description 1
- 238000011056 performance test Methods 0.000 description 1
- 229920000642 polymer Polymers 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Traffic Control Systems (AREA)
- Optical Radar Systems And Details Thereof (AREA)
Abstract
The invention relates to a simulation test system of a ship-borne laser sensing module, which comprises a laser sensing module, a simulated ship motion platform, a platform motion control module, a virtual scene simulation module, a data acquisition module, a data processing module and a comparison result display module, wherein the laser sensing module is arranged on the simulated ship motion platform and is used for sensing models of various obstacles; the simulated ship motion platform is used for simulating the ship motion condition; the platform motion control module is used for controlling the motion condition of the simulated ship motion platform to be the same as the motion condition of the ship under the real sea condition; the virtual scene simulation module is a built model of various obstacles; the data acquisition module is used for acquiring sensing data; the data processing module is used for processing the sensing data and comparing the processing result with the barrier information; and the comparison result display module is used for displaying the comparison result. The simulation test result has high authenticity and low cost.
Description
Technical Field
The invention belongs to the field of ships, relates to a ship simulation test system, and particularly relates to a simulation test system for a ship carrying a laser sensing module.
Background
With the rapid development of the shipping industry, the intelligent development trend of the shipping industry is more and more urgent. The intelligent development of the ship has higher and higher requirements on the intelligent ship. The marine environment is changeable instantly and difficult to predict, the intelligent ship sails in a complex marine environment, and the detection of obstacle information around the sailing environment has great influence on the sailing safety of the ship.
To the perception of intelligent boats and ships navigation surrounding environment, the system of adoption is including multiple such as radar, sonar, vision, but this detection instrument all can receive factors such as distance and environment to disturb, leads to the barrier information that detects imperfect or even wrong. The problems are solved by adopting the laser radar for detection and combining the characteristics of the laser.
However, most of tests on the detection and identification functions of the laser sensing equipment are performance tests performed before shipment, and are conventional means for equipment detection and testing, and the working performance of the sensing equipment under the actual navigation sea condition cannot be tested, so that the safety of a ship cannot be guaranteed.
In addition, the conventional acquisition means for the working performance data of the laser sensing equipment is real ship navigation data, and the data acquired by the part is missing, so that the data acquisition cannot be carried out on the running state of the equipment under the extreme sea condition.
In view of the technical defects in the prior art, a simulation test system for a ship carrying a laser sensing module is urgently needed to be developed.
Disclosure of Invention
The invention aims to overcome the defects in the prior art and provides a simulation test system for a ship-mounted laser sensing module, which can test the sensing capability of a ship-mounted laser sensing device.
In order to achieve the above purpose, the invention provides the following technical scheme:
a simulation test system of a ship carried laser sensing module is characterized by comprising a laser sensing module, a simulated ship motion platform, a platform motion control module, a virtual scene simulation module, a data acquisition module, a data processing module and a comparison result display module, wherein,
the laser sensing module is arranged on the simulated ship motion platform and used for sensing models of various obstacles in the virtual scene simulation module so as to form sensing data;
the ship motion simulation platform is used for simulating the ship motion condition;
the platform motion control module is used for controlling the motion condition of the simulated ship motion platform so that the motion condition of the simulated ship motion platform is the same as the motion condition of a ship under a real sea condition;
the virtual scene simulation module is a model of various obstacles encountered by a ship in the real sailing process, which is built in the simulation room range;
the data acquisition module is used for acquiring sensing data of the laser sensing module;
the data processing module is used for processing the sensing data and comparing a processing result with barrier information;
and the comparison result display module is used for displaying the result obtained by the comparison.
Preferably, the data processing module comprises a preliminary classification sub-module, a clustering sub-module, a rejection sub-module, a feature vector representation sub-module and a classification and drawing sub-module,
the primary classification submodule is used for removing noise points in the perception data based on a filtering algorithm;
the clustering submodule is used for clustering the perception data after noise points are removed;
the eliminating submodule is used for eliminating the clustered classes containing the isolated points;
the feature vector representation submodule is used for representing the class with the feature vector after the class containing the isolated points is removed;
the classification and drawing submodule is used for comparing the feature vector of the class with the feature vector of the obstacle information to obtain the obstacle type corresponding to the perception data.
Preferably, the clustering sub-module specifically clusters the perceptual data from which the noise points are removed by: judging whether the distance between a certain data point in the sensing data and the previous data point adjacent to the certain data point is greater than a threshold value or not, and if not, classifying the certain data point and the previous data point adjacent to the certain data point into the same class; if greater than the threshold, then a data point is classified as another class different from the class to which the immediately preceding data point is classified.
Preferably, wherein the perceptual data consists of n data points, the data points in polar coordinates are from the set D { (r)i,θi) 1,2, n, and defines
Is the distance between the ith data point and its immediately preceding data point, dmaxIs a threshold value, then
Wherein the content of the first and second substances,
in the formula, wherein riThe distance value of the ith data point under polar coordinates is referred to; r isi-1The distance value of the last data point adjacent to the ith data point; thetaiIs the angle value of the ith data point; thetai-1The angle value of the last data point adjacent to the ith data point; Δ φ is the angular resolution of the lidar; τ is thresholdA value constant; sigmarThe actual measurement precision of the laser radar is obtained; delta theta is riAnd ri-1Angle therebetween, Δ θ ═ θi-θi-1。
Preferably, wherein the threshold constant τ is 5.
Preferably, wherein the accuracy σ of the actual measurement of the lidarr＝5.4cm。
Preferably, the feature vector representation submodule is configured to use the feature vector representation to specifically represent the class excluding the class containing the outlier as follows: after eliminating the class containing isolated points, the perception data is composed of a characteristic vector C, wherein C is ═ C1,c2,...,cm]TWherein c isiIs each class after the class containing isolated points is removed, m is the number of classes, ciExpressed as its feature vector: c. Ci＝{ki,d1(i),d2(i),ρstd(i),dmax(i),dstd(i)}，
In the formula, kiIs of the class ciThe number of data points involved; d1(i) Is of the class ciThe first data point in (a) is a distance from the previous class ci-1The distance value of the last data point in (a); d2(i) Is of the class ciThe last data point in (a) is a distance of the latter class ci+1The distance value of the first data point in (a); rhostd(i) Is of the class ciThe distance value ρ of the data point in (1)iStandard deviation of (1), ρiRepresents class ciThe distance value of the data point in (1) in polar coordinates; dmax(i) Is of the class ciDistance of data point in (1) from straight lineiMaximum distance of liIs of the class ciA straight line where a connecting line of a first data point and a last data point is located; dstd(i) Is of the class ciThe data point in (1) deviates from the straight line liThe variance of the distance values of (a).
Preferably, the classification and drawing submodule is configured to input a class represented by a feature vector into a classifier, and compare the feature vector of the class with a feature vector of obstacle information through function calculation in the classifier, so as to obtain a type of an obstacle corresponding to the sensing data.
And finally, the classification and drawing submodule is also used for drawing the class represented by the feature vector.
Compared with the prior art, the simulation test system for the ship carried laser sensing module has the following beneficial technical effects:
1. the motion condition of the ship in the real marine environment is simulated in a mode of building a test platform, the accuracy of the obtained data can be ensured by the data obtained by the semi-physical simulation test, and the test result is convincing to the monitoring of the equipment;
2. by adopting the test method, the test cost can be saved, the test time can be saved, the test work of the ship laser sensing module can be completed in a short time, and the test can be repeatedly carried out;
3. the test method of building the test platform is adopted, so that the danger and economic loss caused by special sea conditions are avoided;
4. the data acquired by the sensing equipment is processed by the identification method, so that the tested result is closer to reality and is more guaranteed.
Drawings
Fig. 1 is a schematic configuration diagram of a simulation test system of a ship laser sensing module according to the present invention.
FIG. 2 is a characteristic diagram of data points displayed by obstacles scanned by the laser sensing module during the course of a vessel's voyage.
FIG. 3 is a flow chart of clustering perceptual data in the present invention.
Detailed Description
The present invention is further described with reference to the following drawings and examples, which are not intended to limit the scope of the present invention.
The laser radar is a radar system which emits laser beams to detect the position, speed and other characteristic quantities of a target, and the working principle of the radar system is to emit the detection laser beams to the target, compare received signals reflected from the target with the emitted signals, and obtain the related information of the target after proper processing.
In the present invention, the reason why the laser is selected as the detection means is: laser self has the characteristics such as the directionality is good, and monochromaticity is good, hi-lite and coherence are good, consequently adopts laser radar to survey, compares in other detection tools, has that the directionality is good, the detection precision is high, the test is far away, detection speed is fast, interference immunity is strong, atmospheric transmission performance is good, have certain penetrability and the disguise advantage such as good to the fog. The application of the laser radar to detect the obstacle information in the sea area in the sea is more and more extensive.
Meanwhile, the laser radar can meet the requirement of rapid and accurate sensing of the surrounding environment of the intelligent ship under high-speed navigation, detect obstacles such as ships, navigation marks and piers, and realize the autonomous obstacle avoidance function of the intelligent ship. The simulation test system of the ship-mounted laser sensing module detects and tests the detection and identification functions of the ship laser radar equipment, makes the platform motion situation the same as the real ship motion situation by controlling the parameters of the platform motion in a mode of building a test platform, installs the laser sensing equipment on the test platform, builds obstacles with different distances and different forms in a test area, and scales the sizes of the obstacles according to the actual size in a certain proportion.
The detection function of the laser sensing module carried by the ship is used for testing and detecting, so that the laser sensing module can accurately detect the barrier information on the air route when the ship navigates in a complex marine environment, and the safety, effectiveness and rapidness of the ship navigation are ensured.
Fig. 1 shows a schematic configuration diagram of a simulation test system of a ship laser sensing module of the invention. As shown in fig. 1, the simulation test system of the laser sensing module carried on the ship of the invention comprises a test platform, a simulation module, a data module and a comparison result display module.
The test platform is a simulated ship motion platform and is used for simulating the motion condition of a real ship.
And the simulated ship platform is provided with a laser sensing module. The laser sensing module is a laser radar. The laser perception module is used for perceiving models of various obstacles in the virtual scene simulation module, and accordingly perception data are formed.
The simulated ship platform is controlled by a platform motion control module. The platform motion control module is used for setting and controlling parameters of the simulated ship platform, so that the motion condition of the simulated ship motion platform is the same as the motion condition of a ship under a real sea condition.
The virtual module is the virtual scene simulation module, and is a model of various obstacles which may be encountered by the ship in the real sailing process and built in the simulation room range. The laser sensing module is used for sensing the models of the various obstacles so as to test whether the models of the various obstacles can be correctly identified.
The data module comprises a data acquisition module and a data processing module connected with the data acquisition module.
The data acquisition module is connected with the laser sensing module and used for acquiring sensing data of the laser sensing module.
The data processing module is used for processing the sensing data and comparing a processing result with the barrier information.
The sensing data acquired by the laser sensing device are points and lines. In the laser data, as shown in fig. 2, the data points returned by the large ship are usually characterized by continuous long and dense straight lines, the data points returned by the pier are usually characterized by short and dense broken lines, and the data points returned by the navigation mark are fewer in number but relatively concentrated. Through the analysis of the laser return data (i.e., the perception data), the number of the data points of the object, the angle range corresponding to the data points, the standard deviation of the distance values of the data points, the curvature of the curve of the data points, and the like can be used as the laser data characteristics of different objects, so that the identification of the obstacle is realized.
Therefore, in the invention, the data processing module comprises a primary classification submodule, a clustering submodule, a rejection submodule, a feature vector representation submodule and a classification and drawing submodule.
And the preliminary classification submodule is used for removing noise in the perception data based on a filtering algorithm.
In the process of data acquisition, due to the influences of laser equipment, fluctuation in ship navigation and the like, the perception data acquired by laser has noise points with different degrees. And the elimination of noise points is beneficial to subsequent clustering, target detection, identification and tracking of laser data. In the invention, invalid data is firstly removed according to the effective scanning range of the laser sensing module during filtering, and then classes with fewer points are deleted according to the number of data points in each class obtained after clustering, so that the barrier classification result is more accurate. The effective distance value distribution of the data collected by the marine laser sensing module is obtained between 100-8000 through the parameters of the laser sensing device and the theoretical knowledge of the corresponding laser.
And the clustering submodule is used for clustering the perception data after the noise points are removed.
Since the sensing data obtained by the laser sensing module is greatly influenced by the environment, the specific number of the cluster point sets (i.e., classes) is difficult to determine. During clustering, a proper class value needs to be determined through continuous iteration, and the complexity of calculation is increased.
However, in the case of the data of the laser radar, if adjacent points are close in physical distance, the probability that the adjacent points are the same object is high. Therefore, in the present invention, simple obstacle segmentation can be achieved by setting a threshold value for the distance between adjacent points.
In the present invention, the clustering submodule clusters the perceptual data after removing the noise point as follows: judging whether the distance between a certain data point in the sensing data and the previous data point adjacent to the certain data point is greater than a threshold value or not, and if not, classifying the certain data point and the previous data point adjacent to the certain data point into the same class; if greater than the threshold, then a data point is classified as another class different from the class to which the immediately preceding data point is classified.
Specifically, as shown in fig. 3, the clustering of the perceptual data after noise point removal specifically includes:
first, the first data point is classified as a first class.
Then, it is determined whether a distance between a second data point adjacent to the first data point and the first data point is greater than a threshold. If not, the second data point is also classified as the first class.
Then, whether the distance between a third data point adjacent to the second data point and the second data point is larger than a threshold value is judged. If not, the third data point is also classified as the first class.
This is done until the distance between a certain data point and its immediately preceding data point is greater than a threshold, e.g., the distance between the eleventh data point adjacent to the tenth data point and the tenth data point is greater than a threshold, and then the eleventh data point is classified as the second class, i.e., a new class.
Next, it is determined whether a distance between a twelfth data point adjacent to the eleventh data point and the eleventh data point is greater than a threshold value. If not, then the twelfth data point is also categorized as the second category.
This is done until the distance between a certain data point and its immediately preceding data point is again greater than the threshold, e.g., the distance between the twentieth data point adjacent to the twentieth data point and the twentieth data point is greater than the threshold, and then the twentieth data point is classified into the third class, i.e., a new class.
This loops until the last data point.
Thereby, clustering of data points in the perception data is completed.
Of course, the above-mentioned polymers are usedClass method, how to determine the threshold is very important. In the present invention, assuming that the sensing data is composed of n points, the data point in polar coordinates may be composed of the set D { (r)i,θi) 1, 2.., n } where r isiThe distance value of the ith data point under polar coordinates is referred to; thetaiIs the angle value of the ith data point in polar coordinates. Definition of
Is the distance value between the ith data point and the adjacent last data point, has
In the formula, Δ φ is the angular resolution of the laser radar, and needs to be determined according to the measurement accuracy of the laser. τ is the correlation threshold constant. SigmarThe actual measurement precision of the laser radar.
dmaxIs a threshold value for determining the distance between the current point and the last adjacent point. Distance between two points
And calculating by the cosine law, and if the calculation result exceeds a threshold value, taking the data point i as a breakpoint.
Experiments show that tau is usually an integer less than 10, and the best value is determined by firstly setting an initial value, then comparing different tau values through experimental data and adjusting. Through experimental study, the targets at a longer distance are difficult to distinguish when τ is 2; when τ is 5, segmentation is accurate for a distant target, and data segmentation effect is good when τ is 5 through actual laser data processing.
Precision sigmarThe selection of the distance is also determined through experiments, in dynamic and static environments, the value of the same distance is measured for N times respectively, so that the standard deviation sigma of the measurement result can be used as a reference value of actual measurement precision, and the calculation of the sigma is represented as:
in the formula aiIs the distance value measured for the ith time; mu is the average value of the distances under N measurements; d is the variance of the measured distance; σ is the standard deviation. The calculation results in both the dynamic and static states show that the dynamic standard deviation and the static standard deviation become gradually larger as the laser distance increases, and the dynamic standard deviation is about 3 times the static standard deviation at the same distance, so σ is selected considering that the obstacle to be detected is in the offshore sea area, exceeding 4000cmr＝5.4cm。
And the eliminating submodule is used for eliminating the clustered classes containing the isolated points.
After the filtering classification and the threshold value adaptive clustering segmentation, the data of the laser radar is divided into a plurality of classes. Due to the fact that the environment of the unmanned ship is complex, some typical non-target obstacles appear in the laser collection range and need to be removed, and therefore the subsequent classification processing pressure is relieved.
Whereas non-target obstacles typically appear as isolated points. Therefore, by removing the class including the isolated point, the class to which the data of the non-target obstacle belongs can be removed.
Generally, the number of the data points in the class containing the isolated points is generally 1 and 2, and according to the judgment condition, the classes with the number of the data points in all the classes being 1-2 are removed to obtain effective classes, so that effective information becomes prominent.
The feature vector representation submodule is used for representing the class after the class containing the isolated points is removed by using the feature vector.
After eliminating the class containing isolated points, the perception data is composed of a feature vector C, and C ═ C1,c2,...,cm]T. Wherein, ciEach class is removed from the class containing isolated points, and m is the number of classes. Class ciThe number of data points contained is ki. Each class ciSet of vectors (p)i,xi,yi) And (4) forming. Rhoi、xiAnd yiRespectively represent class ciThe distance value of the included data point in polar coordinates and the coordinate value converted to rectangular coordinates, i.e. ci＝{(ρi,xi,yi)|i＝1,2,...m}。
Each class ciThe number of data points contained is ki. Class ciThe first point (x) ini,1,yi,1) Class c preceding distancei-1Last point in (1)
A distance value d of1(i) Comprises the following steps:
class ciLast point in (1)
Distance of the latter class ci+1The first point (x) ini+1,1,yi+1,1) Distance betweenOff value d2(i) Comprises the following steps:
in the above formula ki＜19。
Class ciThe distance value ρ of the data point in (1)iStandard deviation of (1) < rho >std(i)：
Class ciIs from the straight line liMaximum distance dmax(i) And l ofiIs of the class ciA straight line where the line connecting the first point and the last point is located, di,jIs of the class ciIs from the straight line liDistance of dmaxExpressed as:
dmax(i)＝max{di,j},j＝2,3,...ki-1
class ciIs deviated from the straight line liD of the distance value ofstd(i) Expressed as:
in the above formula ki＞19。
Then class ciFor its feature vector representation: c. Ci＝{ki,d1(i),d2(i),ρstd(i),dmax(i),dstd(i)}。
The classification and drawing submodule is used for comparing the feature vector of the class with the feature vector of the obstacle information to obtain the obstacle type corresponding to the perception data.
Specifically, the classification and drawing submodule is configured to input a class represented by a feature vector into a classifier, and compare the feature vector of the class with a feature vector of the obstacle information through function calculation in the classifier.
And processing the perception data consisting of the feature vectors C by adopting a target classification technology. The target classification technology means that similar objects have different geometric characteristics, and proper characteristics are selected to form a characteristic vector, wherein the characteristic vector is k in the inventioniClass c thereofiThe number of data points involved; d1(i) Class c thereofiThe first data point in (a) is a distance from the previous class ci-1The distance value of the last data point in (a); d2(i) Class c thereofiThe last data point in (a) is a distance of the latter class ci+1The distance value of the first data point in (a); rhostd(i) Which is of the class ciThe distance value ρ of the data point in (1)iStandard deviation of (1), ρiRepresents class ciThe distance value of the data point in (1) in polar coordinates; dmax(i) Which is of the class ciDistance of data point in (1) from straight lineiMaximum distance of liIs of the class ciA straight line where a connecting line of a first data point and a last data point is located; dstd(i) Which is of the class ciThe data point in (1) deviates from the straight line liThe variance of the distance values of (a).
And inputting the characteristic vector into a classifier, classifying the acquired data through different function calculations, and drawing a classified result. The sample parameters of the classifier comprise clusters c represented by the feature vectors and a classification effect y, namely the sample parameters are represented as: { (c)i,yi)|yiE.g., Y, i is 1,2,.. m, and the clustering c and the classification effect Y of the feature vector are parameters in a classifier, which belongs to the well-known technology. The classification effect is determined according to a classifier function, which is the content of function calculation in the data classifier, and is a known technology without much description. The sample parameters are sample parameters in the classifier, the sample parameters are only obtained by processing the sensing data acquired by the laser radar, then carrying out target classification processing on the processed data, describing results, and judging whether the data are piers, navigation marks and large ships or not according to the depicted results. Wherein m is the same asThe number of the books; y is the category in which the obstacle is identified.
The comparison result display module is used for displaying the result obtained by comparison, so that researchers can visually know the sensing precision and accuracy of the ship laser sensing module.
The simulation test system for the ship-mounted laser sensing module simulates the motion condition of a ship in a real marine environment by adopting a test platform building mode, the data acquired by semi-physical simulation test can ensure the accuracy of the acquired data, and the test result is convincing to the monitoring of laser sensing equipment. Meanwhile, the testing method can save testing cost and testing time, complete the testing work of the ship laser sensing module in a short time and can repeatedly perform testing. Moreover, the test method of building the test platform avoids the danger and economic loss caused by special sea conditions. And finally, the data acquired by the laser sensing module is processed by an identification method, so that the tested result is closer to reality and is more guaranteed. The above examples of the present invention are merely examples for clearly illustrating the present invention and are not intended to limit the embodiments of the present invention. Other variations and modifications will be apparent to persons skilled in the art in light of the above description. Not all embodiments are exhaustive. All obvious changes and modifications which are obvious to the technical scheme of the invention are covered by the protection scope of the invention.
Claims (9)
1. A simulation test system of a ship carried laser sensing module is characterized by comprising a laser sensing module, a simulated ship motion platform, a platform motion control module, a virtual scene simulation module, a data acquisition module, a data processing module and a comparison result display module, wherein,
the laser sensing module is arranged on the simulated ship motion platform and used for sensing models of various obstacles in the virtual scene simulation module so as to form sensing data;
the simulated ship motion platform is used for simulating the motion condition of a real ship;
the platform motion control module is used for controlling the motion condition of the simulated ship motion platform so that the motion condition of the simulated ship motion platform is the same as the motion condition of a ship under a real sea condition;
the virtual scene simulation module is a model of various obstacles encountered by a ship in the real sailing process, which is built in the simulation room range;
the data acquisition module is used for acquiring sensing data of the laser sensing module;
the data processing module is used for processing the sensing data and comparing a processing result with barrier information;
and the comparison result display module is used for displaying the result obtained by the comparison.
2. The simulation test system of the laser sensing module carried on the ship according to claim 1, wherein the data processing module comprises a preliminary classification sub-module, a clustering sub-module, a rejection sub-module, a feature vector representation sub-module and a classification and drawing sub-module,
the primary classification submodule is used for removing noise points in the perception data based on a filtering algorithm;
the clustering submodule is used for clustering the perception data after noise points are removed;
the eliminating submodule is used for eliminating the clustered classes containing the isolated points;
the feature vector representation submodule is used for representing the class with the feature vector after the class containing the isolated points is removed;
the classification and drawing submodule is used for comparing the feature vector of the class with the feature vector of the obstacle information so as to obtain the obstacle type perceived by the perception data.
3. The simulation test system for the laser sensing module carried on the ship according to claim 2, wherein the clustering sub-module is specifically configured to cluster the sensing data without noise: judging whether the distance between a certain data point in the sensing data and the previous data point adjacent to the certain data point is greater than a threshold value or not, and if not, classifying the certain data point and the previous data point adjacent to the certain data point into the same class; if greater than the threshold, then a data point is classified as another class different from the class to which the immediately preceding data point is classified.
4. The simulation test system for the laser sensing module mounted on the ship of claim 3, wherein the sensing data is composed of n data points, and the data points in polar coordinates are set from the group of D { (r)i,θi) 1,2, n, and defines
Is the distance between the ith data point and its immediately preceding data point, dmaxIs a threshold value, then
Wherein the content of the first and second substances,
in the formula, riThe distance value of the ith data point under polar coordinates is referred to; r isi-1The distance value of the previous data point adjacent to the ith data point under the polar coordinate is obtained; thetaiIs the angle value of the ith data point under polar coordinates; thetai-1The angle value of the previous data point adjacent to the ith data point under the polar coordinate is obtained; Δ φ is the angular resolution of the lidar; τ is a threshold constant; sigmarTo activateThe accuracy of actual measurement by the optical radar; delta theta is riAnd ri-1Angle therebetween, Δ θ ═ θi-θi-1。
5. The simulation test system for the ship-mounted laser sensing module according to claim 4, wherein the threshold constant τ is 5.
6. The simulation test system for the ship-mounted laser sensing module according to claim 5, wherein the accuracy σ of actual measurement of the laser radar isr＝5.4cm。
7. The simulation test system of the laser sensing module carried on the ship according to claim 6, wherein the feature vector representation submodule is configured to use the feature vector representation to exclude the class containing the isolated point, and specifically: after eliminating the class containing isolated points, the perception data is composed of a characteristic vector C, wherein C is ═ C1,c2,...,cm]TWherein c isiIs each class after the class containing isolated points is removed, m is the number of classes, ciExpressed as its feature vector: c. Ci＝{ki,d1(i),d2(i),ρstd(i),dmax(i),dstd(i)}，
In the formula, kiIs of the class ciThe number of data points involved; d1(i) Is of the class ciThe first data point in (a) is a distance from the previous class ci-1The distance value of the last data point in (a); d2(i) Is of the class ciThe last data point in (a) is a distance of the latter class ci+1The distance value of the first data point in (a); rhostd(i) Is of the class ciThe distance value ρ of the data point in (1)iStandard deviation of (1), ρiRepresents class ciThe distance value of the data point in (1) in polar coordinates; dmax(i) Is of the class ciDistance of data point in (1) from straight lineiMaximum distance of liIs of the class ciThe first data point and the last data point in (1)The straight line on which the line is located; dstd(i) Is of the class ciThe data point in (1) deviates from the straight line liThe variance of the distance values of (a).
8. The simulation test system of the laser sensing module carried on the ship according to claim 7, wherein the classification and drawing submodule is configured to input a class represented by a feature vector into a classifier, and compare the feature vector of the class with a feature vector of obstacle information through function calculation in the classifier, so as to obtain a type of an obstacle corresponding to the sensing data.
9. The system for simulation test of a ship-mounted laser sensing module according to claim 8, wherein the classification and drawing submodule is further configured to draw a class represented by the feature vector.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110176950.XA CN112883643A (en) | 2021-02-07 | 2021-02-07 | Simulation test system for ship carrying laser sensing module |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110176950.XA CN112883643A (en) | 2021-02-07 | 2021-02-07 | Simulation test system for ship carrying laser sensing module |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112883643A true CN112883643A (en) | 2021-06-01 |
Family
ID=76056302
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110176950.XA Pending CN112883643A (en) | 2021-02-07 | 2021-02-07 | Simulation test system for ship carrying laser sensing module |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112883643A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106407905A (en) * | 2016-08-31 | 2017-02-15 | 电子科技大学 | Machine learning-based wireless sensing motion identification method |
| CN106951498A (en) * | 2017-03-15 | 2017-07-14 | 国信优易数据有限公司 | Text clustering method |
| US20170329804A1 (en) * | 2016-05-10 | 2017-11-16 | Libo Fu | Method And Apparatus Of Generating Image Characteristic Representation Of Query, And Image Search Method And Apparatus |
| CN111240226A (en) * | 2020-01-19 | 2020-06-05 | 智慧航海（青岛）科技有限公司 | Simulation test system and method for shipborne visual perception equipment |
-
2021
- 2021-02-07 CN CN202110176950.XA patent/CN112883643A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170329804A1 (en) * | 2016-05-10 | 2017-11-16 | Libo Fu | Method And Apparatus Of Generating Image Characteristic Representation Of Query, And Image Search Method And Apparatus |
| CN106407905A (en) * | 2016-08-31 | 2017-02-15 | 电子科技大学 | Machine learning-based wireless sensing motion identification method |
| CN106951498A (en) * | 2017-03-15 | 2017-07-14 | 国信优易数据有限公司 | Text clustering method |
| CN111240226A (en) * | 2020-01-19 | 2020-06-05 | 智慧航海（青岛）科技有限公司 | Simulation test system and method for shipborne visual perception equipment |
Non-Patent Citations (1)
| Title |
|---|
| 王贵槐等: "《基于激光雷达的内河无人船障碍物识别方法》", 《光学技术》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105841688B (en) | A kind of ship auxiliary anchors alongside the shore method and system | |
| KR20220155559A (en) | Autonomous navigation method using image segmentation | |
| CN110472500A (en) | A kind of water surface sensation target fast algorithm of detecting based on high speed unmanned boat | |
| KR102466804B1 (en) | Autonomous navigation method using image segmentation | |
| EP3239926A1 (en) | Position measuring equipment | |
| CN109029465A (en) | A kind of unmanned boat tracking and obstacle avoidance system based on millimetre-wave radar | |
| KR102530847B1 (en) | Method and device for monitoring harbor and ship | |
| CN109284663A (en) | A kind of sea obstacle detection method based on normal state and uniform Mixture Distribution Model | |
| CN115077414A (en) | Device and method for measuring bottom contour of sea surface target by underwater vehicle | |
| CN110119671B (en) | Underwater sensing method based on artificial lateral line visual image | |
| KR20240080189A (en) | Distance measurement method and distance measurement device using the same | |
| JP3395105B2 (en) | Ship risk assessment system | |
| CN113933828A (en) | Unmanned ship environment self-adaptive multi-scale target detection method and system | |
| CN112883643A (en) | Simulation test system for ship carrying laser sensing module | |
| CA3237725A1 (en) | Systems and methods for draft calculation | |
| CN114037736B (en) | Vehicle detection and tracking method based on self-adaptive model | |
| CN116964472A (en) | Method for detecting at least one object of an environment by means of a reflected signal of a radar sensor system | |
| CN114994705A (en) | Unmanned ship obstacle avoidance method based on laser radar angle correction | |
| CN117337453A (en) | Ship monitoring system, ship monitoring method, information processing device, and program | |
| Hagen et al. | Exploration of COLREG-relevant parameters from historical AIS-data | |
| JP2022048585A (en) | Isar image data learning device, target classification device, and radar device | |
| JP3589186B2 (en) | Onboard marine forecasting device | |
| CN110895680A (en) | Unmanned ship water surface target detection method based on regional suggestion network | |
| bin Mohamad Rafi et al. | Radar-Based Collision Avoidance on Unmanned Surface Vehicles (USV) | |
| CN117788463B (en) | Ship draft detection method based on video AI and multi-mode data fusion |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210601 |