CN111709187A - Modeling method for radar complex electromagnetic environment effect mechanism - Google Patents
Modeling method for radar complex electromagnetic environment effect mechanism Download PDFInfo
- Publication number
- CN111709187A CN111709187A CN202010552903.6A CN202010552903A CN111709187A CN 111709187 A CN111709187 A CN 111709187A CN 202010552903 A CN202010552903 A CN 202010552903A CN 111709187 A CN111709187 A CN 111709187A
- Authority
- CN
- China
- Prior art keywords
- radar
- bayesian network
- electromagnetic environment
- complex electromagnetic
- effect
- 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
-
- 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
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
- G06F18/24155—Bayesian classification
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Probability & Statistics with Applications (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses a modeling method of a radar complex electromagnetic environment effect mechanism. The method provided by the invention comprises the following steps: firstly, learning a Bayesian network structure by using a K2 algorithm on effect data of each subsystem of a radar in an electromagnetic environment, and establishing a Bayesian network topological structure of the radar system from an input element to an output element; then, under the condition that the network structure is known, a maximum likelihood estimation algorithm is adopted to calculate Bayesian network parameters; then, after a Bayesian network model is obtained, sensitivity analysis based on the change degree of the conditional probability is carried out on the radar electromagnetic environment effect; and finally, carrying out probabilistic reasoning on the model of the radar complex electromagnetic environment effect mechanism by adopting a junction tree reasoning algorithm. The method can effectively model the effect mechanism of the radar in the complex electromagnetic environment, the modeling precision is 91%, and theoretical support is provided for improving the fighting capacity of the radar in the complex electromagnetic environment.
Description
Technical Field
The invention belongs to the field of radar complex electromagnetic environment effect mechanisms, and particularly relates to a modeling method of a radar complex electromagnetic environment effect mechanism.
Background
Modern war is a war that is focused on competing for information rights and electromagnetic space. The development history of electronic countermeasure is closely linked to the development of radio technology, electronic technology applied to military equipment. Meanwhile, with the improvement of the informatization degree, the combat efficiency of the radar is often influenced by the electromagnetic environment of the electronic information system in the electromagnetic battlefield environment. The key to improving the electromagnetic environment adaptability of the electronic information system is to find out the influence process of each element of the electromagnetic environment on the information system, that is, to analyze the radar electromagnetic environment effect mechanism.
Due to the fact that the electronic information system is complex in internal structure and different in characteristics, electromagnetic environment parameters cannot be accurately obtained, key action factors cannot be accurately determined, unknown mechanisms cannot be timely found and cleaned, and the like. The research on the complex electromagnetic environment in China starts late, and the documents which can be studied are few and few, so that the method has a great gap with the advanced level in foreign countries. Therefore, a theoretical method capable of modeling the effect mechanism of the complex electromagnetic environment is needed.
Because the modeling precision of the traditional method is low, and the model complexity corresponding to the radar complex electromagnetic environment effect mechanism is relatively high, the traditional method is generally a multi-connection network, and the reasoning problem of a plurality of query nodes is often involved in the reasoning. The Bayesian network adopted by the invention has strong advantages in reasoning problems containing a plurality of query nodes, can quickly calculate the accurate value of marginal distribution or conditional distribution of the target node, and has higher model accuracy.
Disclosure of Invention
The invention provides a modeling method of a radar complex electromagnetic environment effect mechanism, which aims to solve the problems that the complex electromagnetic environment effect mechanism is in a certain space, the factors are various and difficult to clean, and the uncertainty among the effect factors in the radar complex electromagnetic environment cannot be well processed.
In order to achieve the purpose of the invention, the scheme provided by the invention is as follows: a modeling method of radar complex electromagnetic environment effect mechanism comprises the following steps:
step 2, a Bayesian network structure learning stage: carrying out structure learning of a Bayesian network by adopting a K2 algorithm based on scores to obtain a Bayesian network topology structure;
step 3, Bayesian network parameter learning stage: under the condition that the network structure is known, calculating parameters of the Bayesian network based on a statistical method, and obtaining conditional probability description of a target node;
step 4, radar electromagnetic environment effect sensitivity analysis stage: after a Bayesian network model is obtained, the conditional probability distribution of the nodes is obtained through the model, and then the sensitivity or the contribution degree among effect characteristics is analyzed;
step 5, precision reasoning stage of the Bayesian network model: firstly, conducting Bayesian network structure learning on radar effect data, then utilizing parameters of Bayesian network to learn and obtain a final Bayesian network model of a radar complex electromagnetic environment effect mechanism, after obtaining new radar data, adopting a junction tree inference algorithm of Bayesian network to calculate the probability of occurrence of target nodes, determining possible input 'elements' of the radar complex electromagnetic environment effect through maximum posterior probability, and finally utilizing known conditions to calculate the inference precision of the established model.
In the above method, the step 2 includes the following steps:
step 201, classifying the data set obtained in step 1 into a training set and a test set;
step 202, determining a node sequence according to the flow direction of the radar signal;
and step 203, carrying out structure learning of the Bayesian network by using a K2 algorithm to obtain a Bayesian network topological structure of the radar effect data.
In the above method, the statistical method in step 3 is a maximum likelihood estimation algorithm.
In the above method, the step 4 includes the following steps:
step 401, establishing a sensitivity index based on a Bayesian network;
step 402, obtaining a reasoning result required by sensitivity index calculation by using a direct reasoning algorithm;
and 403, calculating the sensitivity among the nodes.
Compared with the prior art, the invention has the beneficial effects that:
1) the invention fully considers the advantage that signals in a radar system have a certain sequence flow direction, and adopts a K2 algorithm based on a node sequence as a structure learning algorithm of a Bayesian network. The method comprises the steps of assigning corresponding radar signal sequences to nodes of the Bayesian network by using a K2 algorithm, then training by using sample data to obtain structure scores of the Bayesian network, then sequentially obtaining structures of the Bayesian network with different scores by using a greedy algorithm, and finally screening the Bayesian structure with the highest score as a final Bayesian network topological structure. The method can effectively model the effect mechanism of the radar complex electromagnetic environment, so that the Bayesian network topology structure is more accurate and can reach 91%, and theoretical support is provided for improving the fighting capacity of the radar in the complex electromagnetic environment.
2) The modeling method is considered to be finally applied to practical application, so that the maximum likelihood estimation method is selected to calculate parameters in the parameter learning of the Bayesian network, the iteration is simple, the difference between the result and the Bayesian estimation is few, and the problem of low model training speed when the data volume is large is solved. The invention can efficiently calculate the parameters of the established model structure, thereby greatly improving the calculation speed under the condition of large sample data.
3) Aiming at the problem of complex uncertain relation which is difficult to clear among radar effect characteristics, the sensitivity analysis method based on the conditional probability change degree is provided, the contribution degree relation among the effect characteristics is further analyzed, and the relation description among the radar effect elements can be more specific.
Drawings
FIG. 1 is a flow chart of a method of an embodiment of the present invention;
FIG. 2 is a schematic representation of a Bayesian network for sensitivity analysis;
FIG. 3 is a flow chart of a junction tree algorithm;
FIG. 4 is a graph of Bayesian network structure learning results;
FIG. 5 is a graph of Bayesian network parameter learning results;
FIG. 6 sensitivity analysis;
FIG. 7 is a diagram of Bayesian network inference accuracy results;
Detailed Description
The present invention will be described in further detail with reference to the following detailed description and accompanying drawings.
The invention provides a modeling method of a radar complex electromagnetic environment effect mechanism, which comprises the following steps:
Said data comprising: the data of each subsystem such as the radar complex electromagnetic environment, the radar receiving front end, the radar signal processing, the radar data processing and the like in the radar system in the real environment are collected, so that a data set of the radar complex electromagnetic environment effect mechanism is obtained.
Step 2, a structure learning stage of the Bayesian network: the radar effect data set is classified and preprocessed, and the data set is divided into two parts according to the proportion, namely a training set and a testing set, so that better model training and model verification effects are provided. The method comprises the following specific steps: step 201, classifying the acquired data set into a training set and a test set; step 202, determining a node sequence according to the flow direction of the radar signal; and step 203, carrying out structure learning of the Bayesian network by using a K2 algorithm to obtain a Bayesian network topological structure of the radar data. The detailed description is as follows:
firstly, in an undirected graph without any edge, only the sequence of all nodes is given, and in the searching process, an algorithm searches a father node set of each node from a first node according to the given node sequence, so that the Bayesian score value of a local network structure formed by the node and the father node is increased continuously, and the searching is stopped until the Bayesian score value is not increased any more or the set upper limit of the number of the father nodes of each node is reached.
Step 3, a parameter learning stage of the Bayesian network: under the condition that the topological structure of the Bayesian network of the radar data sample is known, the conditional probability distribution of each node in the Bayesian network is calculated by utilizing a maximum likelihood estimation algorithm, so that the quantitative representation of the influence relation among the nodes is obtained. The detailed description is as follows: firstly, a likelihood function about a sample is solved; then, taking logarithm of the likelihood function; then, the derivation is carried out on the likelihood function after the logarithm is taken; and finally, making the derivative of the likelihood function be 0, solving the likelihood equation, and obtaining the maximum parameter theta meeting the equation as the solved parameter.
Step 4, radar electromagnetic environment effect sensitivity analysis stage: through the established Bayesian network model, probability description of the magnitude of the influence relation among the effect characteristics can be obtained, then the degree of change of the target effect characteristics along with the change of other related effect characteristics is calculated based on the conditional probability distribution to measure the contribution of each related effect variable to the target effect variable, and further the sensitivity analysis of the radar electromagnetic environment effect based on the Bayesian network is completed. The method comprises the following specific steps: step 401, establishing a sensitivity index based on a Bayesian network; step 402, obtaining a reasoning result required by sensitivity index calculation by using a direct reasoning algorithm; and 403, calculating the sensitivity among the nodes.
Step 5, precision reasoning stage of the Bayesian network model: firstly, carrying out Bayesian network structure learning on sample data; secondly, learning and acquiring a final Bayesian network model of the radar complex electromagnetic environment effect mechanism by utilizing parameters of the Bayesian network; and finally, after new radar data are obtained, calculating the probability of occurrence of the target node by using a junction tree inference algorithm of the Bayesian network, and calculating the inference precision of the established model.
The detailed description is as follows:
the precision inference phase of the Bayesian network model comprises two parts: modeling learning and reasoning. Bayesian network learning is to find a Bayesian model which can truly reflect the mutual dependency relationship among the existing research objects in the current research problem. The Bayesian network learning can be divided into two stages of (i) structure learning, namely learning of a network topology. And secondly, parameter learning, namely learning of local prior conditional probability distribution of each node variable in the network. The Bayesian network inference estimates the value of an unknown node from some nodes of known given values. Under the condition of a given Bayesian network model, the probability of the target node of interest is calculated by utilizing a conditional probability calculation method in Bayesian probability according to known conditions.
Example (b):
as shown in fig. 1, a modeling method of a radar complex electromagnetic environment effect mechanism is implemented by the following specific steps:
s 1: in an actual scene, the representation of the radar complex electromagnetic environment and various measurement technologies are applied to measure the effect data of subsystems such as the radar complex electromagnetic environment, a radar receiving front end, radar signal processing, radar digital processing and the like of a radar system, and data set acquisition is carried out.
s 2: the data sets obtained after actual measurement acquisition are classified in proportion into a training set and a testing set for better model training and model verification effects. The method comprises the following specific steps:
(1): classifying the acquired data set into a training set and a test set;
(2): determining a node order according to the flow direction of the radar signals;
(3): and (5) carrying out structure learning of the Bayesian network by using a K2 algorithm to obtain a Bayesian network topological structure of the radar data.
s 3: and (3) obtaining a Bayesian network topological structure after learning the Bayesian network structure in the step (2), and then calculating parameters of the Bayesian network by utilizing maximum likelihood estimation. The basic idea of maximum likelihood estimation is: d may be present in a random trial1,D2......,DnResults of various tests, assuming in one test, DmIf it occurs, D can be consideredmThe probability of occurrence is the greatest, and the parameter value θ at which the likelihood function P (C | θ) takes the maximum value is taken as the estimated value for the parameter. Possibility to generate samples D from theta, i.e. likelihood function
The bayesian network local likelihood function can be further decomposed into:
in a parent node set
When the information is known, the information is transmitted to the mobile terminal,
distribution is as follows
Other values of
Independent polynomial distribution irrelevant, maximum likelihood estimation method can be calculated
Obtaining the estimated parameters, and using the formula, the method can be usedThe parameters of a given structure are readily available. Therefore, quantitative representation of the magnitude of the influence relationship among the nodes is obtained.
And s4, obtaining the Bayesian network model of the radar complex electromagnetic environment effect data through the steps 2 and 3, and then analyzing the contribution degree between effect nodes by using a sensitivity analysis method based on conditional probability distribution, namely the degree of the change of the target node along with the change of other effect nodes. The method comprises the following specific steps:
(1) providing a sensitivity index based on the Bayesian network;
(2) acquiring a reasoning result required by sensitivity index calculation by using a direct reasoning algorithm;
(3) sensitivity between nodes is calculated.
Specifically, assume that fig. 2 is a bayesian network of the radar effect mechanism that we have established, where Y represents the target effect node and X represents other relevant effect feature nodes. The subject proposes a sensitivity index calculation method based on the degree of change of conditional probability, using X1For example, as shown in equation (3). The mathematical meaning of equation (3) is: and measuring the contribution degree of the related effect characteristics to the target effect characteristics by using the degree that the conditional probability of the target effect variables changes along with the change of the values of other related effect variables.
In the formula:
number of sampling states of q-Y
rx1——XiNumber of value state of
Through parameter learning, P (X) can be obtained1)，P(Xi|π(Xi) And are) and
where i is 2, …, n. The conditional probability formula and the direct reasoning algorithm can be used to know that:
the elements X can be obtained by substituting the formulas (4) and (5) into the formula (3)1Degree of contribution to effect Y
S5: and (3) obtaining a Bayesian network model of the sample data after the Bayesian network structure learning and the parameter learning in the step (2) and (3), and then calculating the probability of the target node by using a junction tree algorithm. As shown in fig. 3, the basic flow of the junction tree inference algorithm: and transforming the Bayesian network into a joint tree, initializing the joint tree according to the probability distribution of the Bayesian network, setting evidence, selecting a cluster as a root node, transmitting messages under the condition of meeting the global consistency requirement, and finishing reasoning.
The embodiment of the invention can model and reason the effect mechanism of the radar complex electromagnetic environment, and carries out simulation modeling test by utilizing Matlab environment under the existing condition of actually measuring radar data samples. The invention uses 20000 pieces of data to show the processes of structure learning, parameter learning and sensitivity analysis, and the rest 2000 pieces of data are used as a verification set, thereby reasoning the precision of the Bayesian network model.
The first step is as follows: and loading data, and selecting the loaded file by self.
The second step is that: after loading the data, structure learning (topology learning of the bayesian network according to the K2 algorithm) is performed, as shown in fig. 4. The Bayesian network is a directed arc segment consisting of nodes and directed arc segments, the nodes represent effect characteristic variables, the arc segments represent causal relations or probability relations, and the arc segments are directed and do not form a loop.
The third step: parameter learning, a conditional probability table of the entire network node is acquired on the basis of the structure obtained by the structure integration, as shown in fig. 5. This is shown as a conditional probability table for node 1, i.e., a quantitative description of the dependencies affecting node 1.
The fourth step: and (3) sensitivity analysis, namely loading the learning result of the model and analyzing the sensitivity or contribution degree between the effect characteristics, as shown in figure 6.
The fourth step: the process of inputting the learning result of the bayesian network and calculating the inference accuracy is shown in fig. 7.
Influence relations among all nodes can be calculated through sensitivity analysis, and taking fig. 6 as an example, it can be seen that a node 5 (false target interval of deception jamming) is related to a columnar convex node in the graph, wherein influence relations of some nodes are larger, such as the node 6 (clutter Doppler spread), 7 (clutter center frequency), 8 (background signal power) and the like; there are small effects such as node 10 (second-stage mixer spectral mean) and node 26 (false track rate), and there are some nodes that have no effect on node 5, such as nodes 1, 2, 3, 11, etc.
The method can effectively model radar effect data through the modeling simulation experiment, can calculate the probability of occurrence of unknown radar effect nodes and the reasoning precision of a model by utilizing the advantages of solving the uncertainty problem through the Bayesian network, and provides powerful support for improving the combat performance of the radar in the complex electromagnetic environment.
The present invention has been described in terms of specific examples, which are provided to aid understanding of the invention and are not intended to be limiting. For a person skilled in the art to which the invention pertains, several simple deductions, modifications or substitutions may be made according to the idea of the invention.
Claims (4)
1. A modeling method of radar complex electromagnetic environment effect mechanism comprises the following steps:
step 1, data set acquisition stage: acquiring effect data of each subsystem of the radar system through actual measurement acquisition;
step 2, a Bayesian network structure learning stage: carrying out structure learning of a Bayesian network by adopting a K2 algorithm based on scores to obtain a Bayesian network topology structure;
step 3, Bayesian network parameter learning stage: under the condition that the network structure is known, calculating parameters of the Bayesian network based on a statistical method, and obtaining conditional probability description of a target node;
step 4, radar electromagnetic environment effect sensitivity analysis stage: after a Bayesian network model is obtained, the conditional probability distribution of the nodes is obtained through the model, and then the sensitivity or the contribution degree among effect characteristics is analyzed;
step 5, precision reasoning stage of the Bayesian network model: firstly, conducting Bayesian network structure learning on radar effect data, then utilizing parameters of Bayesian network to learn and obtain a final Bayesian network model of a radar complex electromagnetic environment effect mechanism, after obtaining new radar data, adopting a junction tree inference algorithm of Bayesian network to calculate the probability of occurrence of target nodes, determining possible input 'elements' of the radar complex electromagnetic environment effect through maximum posterior probability, and finally utilizing known conditions to calculate the inference precision of the established model.
2. The modeling method of radar complex electromagnetic environment effect mechanism according to claim 2, wherein the step 2 comprises the following steps:
step 201, classifying the data set obtained in step 1 into a training set and a test set;
step 202, determining a node sequence according to the flow direction of the radar signal;
and step 203, carrying out structure learning of the Bayesian network by using a K2 algorithm to obtain a Bayesian network topological structure of the radar effect data.
3. The modeling method of radar complex electromagnetic environment effect mechanism according to claim 1 or 2, wherein the statistical method in the step 3 is a maximum likelihood estimation algorithm.
4. The modeling method of radar complex electromagnetic environment effect mechanism according to claim 3, wherein the step 4 comprises the following steps:
step 401, establishing a sensitivity index based on a Bayesian network;
step 402, obtaining a reasoning result required by sensitivity index calculation by using a direct reasoning algorithm;
and 403, calculating the sensitivity among the nodes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010552903.6A CN111709187B (en) | 2020-06-17 | 2020-06-17 | Modeling method of radar complex electromagnetic environment effect mechanism |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010552903.6A CN111709187B (en) | 2020-06-17 | 2020-06-17 | Modeling method of radar complex electromagnetic environment effect mechanism |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111709187A true CN111709187A (en) | 2020-09-25 |
| CN111709187B CN111709187B (en) | 2023-06-16 |
Family
ID=72540754
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010552903.6A Active CN111709187B (en) | 2020-06-17 | 2020-06-17 | Modeling method of radar complex electromagnetic environment effect mechanism |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111709187B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030115325A1 (en) * | 2001-12-18 | 2003-06-19 | Ira Cohen | Adapting Bayesian network parameters on-line in a dynamic environment |
| CN106960069A (en) * | 2016-12-27 | 2017-07-18 | 安徽理工大学 | A kind of Bayesian network platform with self-learning function |
| US20190087294A1 (en) * | 2018-05-11 | 2019-03-21 | Beihang University | Method for establishing fault diagnosis technique based on contingent Bayesian networks |
-
2020
- 2020-06-17 CN CN202010552903.6A patent/CN111709187B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030115325A1 (en) * | 2001-12-18 | 2003-06-19 | Ira Cohen | Adapting Bayesian network parameters on-line in a dynamic environment |
| CN106960069A (en) * | 2016-12-27 | 2017-07-18 | 安徽理工大学 | A kind of Bayesian network platform with self-learning function |
| US20190087294A1 (en) * | 2018-05-11 | 2019-03-21 | Beihang University | Method for establishing fault diagnosis technique based on contingent Bayesian networks |
Non-Patent Citations (2)
| Title |
|---|
| 刘春生等: "基于Netica的机载雷达侦察效能评估", 《现代雷达》 * |
| 陈静等: "基于Netica的自学习贝叶斯网络的构建", 《电子测量与仪器学报》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111709187B (en) | 2023-06-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Candy | Bootstrap particle filtering | |
| Li et al. | Missing traffic data: comparison of imputation methods | |
| Raghu et al. | Evaluation of causal structure learning methods on mixed data types | |
| CN110632572A (en) | Radar radiation source individual identification method and device based on unintentional phase modulation characteristics | |
| Bulakh et al. | Time series classification based on fractal properties | |
| Dawid | Prequential data analysis | |
| Lee et al. | Interpretable tropical cyclone intensity estimation using Dvorak-inspired machine learning techniques | |
| Kotenko et al. | An approach for intelligent evaluation of the state of complex autonomous objects based on the wavelet analysis | |
| Hou et al. | A new Bayesian scheme for self-starting process mean monitoring | |
| Bocquel et al. | Multitarget tracking with multiscan knowledge exploitation using sequential MCMC sampling | |
| CN104821854A (en) | Multidimensional spectrum sensing method for multiple main users based on random sets | |
| CN108647186B (en) | Small sample test data estimation method based on grey self-help and unknown rational number | |
| CN111709187A (en) | Modeling method for radar complex electromagnetic environment effect mechanism | |
| Tolba et al. | Bayesian and non-Bayesian estimation methods to independent competing risks models with type II half logistic weibull sub-distributions with application to an automatic life test | |
| Dudczyk et al. | Applying the distance and similarity functions to radar signals identification | |
| Li et al. | Few-shot scenario testing for autonomous vehicles based on neighborhood coverage and similarity | |
| Govalkar et al. | Siamese Network based Pulse and Signal Attribute Identification | |
| Yongguang et al. | Grey distance information approach for parameter estimation of small samples | |
| Pang et al. | Detecting continual anomalies in monitoring data stream based on sampling GPR algorithm | |
| Peng et al. | An improved adaptive ET-PHD algorithm for newborn target intensity | |
| CN115166453B (en) | Partial discharge continuous monitoring method and device based on edge real-time radio frequency pulse classification | |
| Bocquel et al. | Multitarget tracking with IP reversible jump MCMC-PF | |
| Žák et al. | A histogram based radar signal identification process | |
| CN118070639A (en) | Transient electromagnetic emission characteristic probability assessment method | |
| Moratuwage et al. | A moving window based approach to multi-scan multi-target tracking |
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 | ||
| CB03 | Change of inventor or designer information |
Inventor after: Wang Peng Inventor after: He Maimai Inventor after: Di Ruohai Inventor after: Dong Mianmian Inventor after: Wu Jiao Inventor after: Hong Xianli Inventor after: Li Xiaoyan Inventor after: Lv Zhigang Inventor before: Dong Mianmian Inventor before: He Maimai Inventor before: Di Ruohai Inventor before: Wang Peng Inventor before: Wu Jiao Inventor before: Hong Xianli Inventor before: Li Xiaoyan Inventor before: Lv Zhigang |
|
| CB03 | Change of inventor or designer information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |