CN114896144A - Knowledge base-based CPS model equivalent modulus input variation test method - Google Patents
Knowledge base-based CPS model equivalent modulus input variation test method Download PDFInfo
- Publication number
- CN114896144A CN114896144A CN202210451298.2A CN202210451298A CN114896144A CN 114896144 A CN114896144 A CN 114896144A CN 202210451298 A CN202210451298 A CN 202210451298A CN 114896144 A CN114896144 A CN 114896144A
- Authority
- CN
- China
- Prior art keywords
- test case
- zombie
- modules
- module
- areas
- 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
- 238000010998 test method Methods 0.000 title claims description 4
- 238000012360 testing method Methods 0.000 claims abstract description 108
- 238000013136 deep learning model Methods 0.000 claims abstract description 36
- 238000000034 method Methods 0.000 claims abstract description 26
- 230000035772 mutation Effects 0.000 claims abstract description 20
- 230000007246 mechanism Effects 0.000 claims description 22
- 238000005070 sampling Methods 0.000 claims description 11
- 230000008569 process Effects 0.000 claims description 10
- 238000012549 training Methods 0.000 claims description 9
- 238000006243 chemical reaction Methods 0.000 claims description 4
- 238000013507 mapping Methods 0.000 claims description 4
- 230000007274 generation of a signal involved in cell-cell signaling Effects 0.000 claims 1
- 239000011159 matrix material Substances 0.000 description 6
- 230000006870 function Effects 0.000 description 4
- 238000007781 pre-processing Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 239000003550 marker Substances 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 125000000205 L-threonino group Chemical group [H]OC(=O)[C@@]([H])(N([H])[*])[C@](C([H])([H])[H])([H])O[H] 0.000 description 1
- 238000012952 Resampling Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000013522 software testing Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/36—Preventing errors by testing or debugging software
- G06F11/3668—Software testing
- G06F11/3672—Test management
- G06F11/3684—Test management for test design, e.g. generating new test cases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/36—Preventing errors by testing or debugging software
- G06F11/3668—Software testing
- G06F11/3672—Test management
- G06F11/3688—Test management for test execution, e.g. scheduling of test suites
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Quality & Reliability (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- Debugging And Monitoring (AREA)
Abstract
The invention discloses a CPS model equivalent modulus input variation testing method based on a knowledge base, which comprises the following steps: marking the step-by-step areas of all kinds of modules in the test case by using Simulink software; establishing a deep learning model; when the test case is mutated, basic mutation and mutation operation are included; testing the Simulink software, carrying out differential test on the variant test case and the original Simulink test case, compiling and running the test case, obtaining the input and output values of each module in the test case, comparing all the input and output values of the variant test case and the original Simulink test case, and recording the difference information if the difference is generated. The method increases the diversity and randomness of the variant test cases; the problem that the difference between a variant test case and an original Simulink test case is too small is solved by a method of adding a large number of test case modules in a nested zombie region.
Description
Technical Field
The invention relates to the field of software testing, in particular to a CPS model equivalent modulus input variation testing method based on a knowledge base.
Background
The CPS (information physical system) is an intelligent system integrating calculation, communication and control as a whole, wherein the CPS is used as a whole of a calculation process and a physical process. Simulink is a multi-domain simulation in the field of CPS and a model-based design tool. It supports system level design, simulation, automatic code generation, and continuous testing and verification of embedded systems.
Testing CPS development tools is very important because CPS is a de facto industry standard in key fields such as smart grids, autonomous driving automotive systems, medical monitoring, industrial control systems, autonomous driving avionics, and the like. However, the whole Simulink tool chain is expensive to verify formally, so that the current technologies mainly include generating test case operation detection Bug by a random model, detecting Bug by using a differential test, and the like.
Bug for tool chain was tested by a novel mutation approach against Simulink development tool. The prior art has certain defects in the Bug detection process. The specific defects are as follows: variants with insufficient variability: both techniques for detecting Simulink toolchain Bug suffer from certain drawbacks. Testing requires existing test cases and test predicates as benchmarks. But the test predictions are difficult to obtain and too costly. The problem can be alleviated to a certain extent by utilizing EMI variation, but the variation amplitude of the existing EMI variation method is too small at present, and only partial modules in zombie areas are deleted, and blocks are added or replaced on the premise of ensuring data types or output consistency, so that the variation difference is insufficient. The smaller range of variant operations results in the difference test possibly never detecting the Bug of some hidden components.
The variant has insufficient reliability: the invention hopes to add a large amount of content-rich variants in the mutation process, and the realization of the idea is derived from the random generation of a deep learning model. Since the Simulink language does not have a published formal language specification, deep learning model generation can only learn a partial specification of the Simulink model from the seed model and randomly generate from scratch. Sequences generated in sequence from scratch may be subject to large deviations. The RNN deep learning model generated based on time sequence can not well deal with the problem of long dependence of Simulink model text, so that the reliability of the conventional model is relatively poor.
Disclosure of Invention
According to the problems existing in the prior art, the invention discloses a CPS model equivalent modulus input variation testing method based on a knowledge base, which specifically comprises the following steps:
marking the step-by-step areas of all kinds of modules in the test case by using Simulink software;
establishing a deep learning model, wherein the deep learning model comprises an encoder, a decoder and a sampling mechanism, taking a correctly-operated test case as a seed model, inputting the preprocessed and coded seed model into the deep learning model, and training by using the back propagation characteristic of the deep learning model, wherein the deep learning model after training has the capability of generating an effective test case;
when the test case is mutated, the method comprises basic mutation and mutation operation, wherein when the basic mutation is: the data types of all modules of the test case and the potential conflict problem of the sample time are perfected, and then the generation operation is carried out; acquiring nested zombie areas and non-zombie areas of the test cases, randomly selecting a plurality of module sections of the non-zombie areas, inputting the module sections into a trained deep learning model, continuously writing the module sections to generate a Simulink test case capable of running correctly, and independently storing the generated test case in a knowledge base, wherein the number of the test cases in the knowledge base is increased along with the increase of variation times; during mutation operation: traversing the nested zombie areas, deleting partial modules in the nested zombie areas divided by taking the subsystem as a unit, adding test cases randomly selected from a knowledge base in the subsystem, and traversing all the areas to obtain variant test cases;
testing of Simulink software: setting quantitative variation times, carrying out differential test on the variant test case and the original Simulink test case, compiling and running the test case, obtaining the input and output values of each module in the test case, comparing all the input and output values of the variant test case and the original Simulink test case, if the difference is generated, recording the difference information, otherwise, completing the test process.
When the step-by-step areas are obtained, the execution coverage rates of all modules are firstly obtained, so that the variant modules meeting the conditions are judged and defined as zombie blocks, the areas where the zombie blocks are located are used as variant areas input by equivalent modules, the hierarchical structures of the modules are classified, the zombie blocks located at the top layer are removed to obtain middle area zombie blocks, and the middle areas are used as nested zombie areas.
Further, a deep learning model is built by adopting a python language and a Tensorflow framework.
Further, when the seed model is preprocessed: acquiring a text sequence of a test case, deleting default configuration information, default parameter sequence, blank space, annotation and position information in the text sequence, and then interleaving module and connecting line information; when the seed model is coded: and mapping words of common keywords and parameter names to numbers, and mapping each character of the rest text sequences to the numbers respectively to finally obtain the coded text sequences represented by the numbers.
When the potential conflict problem of the data type and the sample time of the module is perfected: adding a type conversion module in front of each module to annotate the type of each module, thereby preventing mutation from changing the data type inference result of the module; the sample time parameters of the signal-generating module are labeled in order for the variation to satisfy an equivalent condition.
When the module segment is subjected to continuous writing to generate the Simulink test case which can run correctly: and acquiring the partial module, writing the partial module into a new file, acquiring a module segment in an mdl format through the file storage characteristic of Simulink software, reading a text sequence of the mdl file, inputting the text sequence into a deep learning model for completing training, and generating a new test case by using a sampling mechanism of the deep learning model and independently storing the new test case in a knowledge base.
When carrying out mutation operation: traversing each nested zombie region, randomly selecting zombie blocks in the region, judging whether the zombie blocks are IF condition blocks or subsystem modules, not deleting, randomly deleting other nested zombie region blocks, storing the front and back connection information of the deleted modules, randomly loading the models in the knowledge base, adding all modules and connecting lines of the whole model in the region, and connecting the head and tail modules of the models to the front and back connection positions of the deleted modules; and repeating the operations until all the nested zombie modules are processed, and obtaining the variant test case.
When carrying out mutation operation: traversing each nested zombie region, randomly selecting zombie blocks in the region, judging whether the zombie blocks are IF condition blocks or subsystem modules, deleting none of the modules, randomly deleting other nested zombie region blocks, and storing the front and back connection information of the deleted modules; randomly loading a model in a knowledge base, adding all modules and connecting lines of the whole model in the region, and connecting the head and the tail of the model with the front module and the rear module of the deleted module; and repeating the operations until all the nested zombie modules are processed, and obtaining the variant test case.
Due to the adoption of the technical scheme, the CPS model equivalent mode input variation test method based on the knowledge base, provided by the invention, has the advantages that the random test case is obtained through deep learning and is used as potential variable knowledge, and the diversity and the randomness of the variant test case are increased; the problem that the difference between the variant test case and the original Simulink test case is too small is solved by a method of adding a large number of test case modules in the nested zombie region, the difference of the variant is expanded, and more reliable and diversified variant test cases are finally obtained, so that the possibility of detecting CPS software Bug by differential test is improved.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments described in the present application, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is a schematic diagram illustrating a mutation process performed on a test case according to the present invention.
Detailed Description
In order to make the technical solutions and advantages of the present invention clearer, the following describes the technical solutions in the embodiments of the present invention clearly and completely with reference to the drawings in the embodiments of the present invention:
as shown in fig. 1, a CPS model equivalent modulus input variation testing method based on a knowledge base specifically includes the following steps:
s1, acquiring nested zombie areas: the fact that the zombie blocks are distinguished from the common modules is determined through whether the modules influence final results in the execution process of the data stream, and the number and the area of the zombie blocks in all the modules of a test case are determined by adopting the execution coverage rate function of Simulink software. Modules except the zombie blocks are not processed temporarily, the test cases have the hierarchical structure of subsystems and nested subsystems, and the zombie blocks are classified according to the characteristic. And eliminating the zombie blocks positioned at the top layer to obtain middle region zombie blocks, and taking the middle region as a nested zombie region.
S2, establishing and training a deep learning model: the method comprises the following three steps:
s21, establishing a deep learning model: the method comprises the steps of establishing a deep learning model by using a python language and a Tensorflow framework, and adopting a classic transducer deep learning model, wherein the structure of the deep learning model comprises an encoder, a decoder and a sampling mechanism.
The encoder is formed by stacking N layers of same self-attention mechanism layers, wherein each self-attention mechanism layer is composed of a position encoding mechanism, a multi-head attention mechanism and a fully-connected feedforward network. The position coding mechanism uses a relative position representation to model mdl text sequence X ═ X 1 ,…,x l ,…,x n The pairwise relationship between. The position-coding formula is as follows:
where pos represents the position of the marker in X, thres represents the threshold, I is the index of each marker in the vector, d model Is the dimension of the vector. The multi-head attention mechanism allows the model to train the self-attention mechanism in parallel, enabling the model to focus on different aspects of the mdl text sequence. Subsequently, a plurality of self-attentive mechanical layers are sequentially connected together, as shown in the following equation:
Z＝MultiHead(Q,K,V)＝Concat(head 1 ...head h )·W o
w herein X A weight matrix representing the conversion of the input into a vector, Q, K and V represent the query vector, the key vector and the value vector, respectively, of the self-attention mechanism. In addition, each W i Are all matrices of Q, K, V linear transformation, and W o A weight matrix representing a multi-headed attention mechanism. The output information of the multi-head attention mechanism is then adjusted by the residual connection and normalization network layer and the feedforward neural network layer as the final output S of the encoder. These equations can be expressed as follows:
L＝LayerNorm(X+Z)
FFN(L)＝max(0,L·W 1 +b 1 )·W 2 +b 2
S＝LayerNorm(L+FFN(L))
wherein, W 1 And W 2 Is the weight matrix of the encoder, LayerNorm is a normalization function, b 1 And b 2 Is the bias matrix of the encoder.
The decoder decodes the hidden layer vector obtained during encoding. The decoder consists of a position coding mechanism, a multi-head attention mechanism, a fully-connected feedforward network and a softmax layer. A position coding mechanism of a decoder, a multi-head attention mechanism and a fully-connected feedforward network and an encoder. Then, the probability of the next code mark in the current state is calculated based on the softmax layer, and the formula is as follows:
wherein W f Is the weight matrix of the full connection layer, S de Is the full connection layer output of the decoder,
is the bias matrix of the decoder.
The sampling mechanism is set as follows: and selecting a text sequence as a seed sequence to be input into the deep learning model, sampling a next word based on the text sequence by the deep learning model, and connecting the seed sequence and the sampled word to be used as the input of the next deep learning model. The next word is sampled in this manner. Meanwhile, some limits are set to stop the sampling of the sequence, for example, (1) stop when the total number of words of the text sequence exceeds 5000; (2) the generation is stopped in case the brackets generated comply with the structural specification. And checking whether the complete sequence obtained by sampling is a valid test case, if so, taking the complete sequence as a result of the sampling mechanism, and if not, resampling.
S22, preprocessing and coding the test case: we obtain the text sequence of the mdl format of the test case, which is directly generated by Simulink software as the format for saving the test case. The test case preprocessing and encoding process comprises the following detailed steps: (1) deleting BlockDefaults {. and AntotationDefaults {. A. }; (2) deleting all other model parameters except the System, such as configuration defaults, graphical interface defaults, to ensure that the model can be compiled; (3) then blank deletion is carried out, long block names are converted into short block names, and any comments and position information are deleted; (4) in order to make the information connection between the blocks and the lines in the generated model more compact, the information of the blocks and the connecting lines is staggered in the mdl file, so that each pair of connected blocks can obtain the connection information of the blocks after being defined; (5) the coding adopts a mixed coding system, common keywords and parameter names are mapped to words, and the rest are mapped to characters.
S23 training the deep learning model
Inputting the test cases after preprocessing and coding into the deep learning model, training the parameters of the deep learning model by using the back propagation capability of the Tensorflow frame, and finally storing the trained parameters of the deep learning model after a fixed-length training round.
As shown in FIG. 2, S3, the test case is mutated:
the basic mutation of the test case comprises two steps: (1) adding a type conversion module in front of each module to annotate the module type of each module, so as to prevent mutation from changing the data type inference result of the module; (2) the sample time parameters of the module generating the signal are labeled in order for the variation to satisfy an equivalent condition.
After basic variation, acquiring nested zombie areas and non-zombie areas in the test case, randomly selecting a part of module segments of the non-zombie areas, inputting the part of module segments into a trained deep learning model, and continuously writing the module segments to generate the Simulink test case capable of running correctly. The detailed steps are as follows: (1) acquiring a new file written in by the partial module; (2) obtaining a module segment in mdl format through a file storage function of Simulink software; (3) and reading the text sequence of the mdl file, inputting the trained deep learning model, generating a new test case by using a sampling mechanism of the deep learning model, and independently storing the new test case in a knowledge base.
The specific operation for the mutation of the nested zombie region is as follows: (1) traversing each nested zombie region, randomly selecting zombie blocks in the region, judging whether the zombie blocks are IF condition blocks or subsystem modules, deleting none of the modules, randomly deleting other nested zombie region blocks, and storing the front and back connection information of the deleted modules; (2) randomly loading a model in a knowledge base, adding all modules and connecting lines of the whole model in the region, and connecting the head and the tail of the model with the front module and the rear module of the deleted module; (3) and repeating the operations until all the nested zombie modules are processed, and obtaining the variant test case.
S4 testing Simulink software:
and setting quantitative variation times, wherein the variation times are determined according to time and the computing power of the server. And carrying out differential test on the variant test case and the original Simulink test case. The differential test requires attention to: (1) under the same server environment, adopting the same default setting, compiling and running the test case by using the compiling and executing functions of Simulink software; (2) during the test process, the output signals of the two test cases are found to generate different results, namely, the Bug is considered to be Bug of a tool chain. More formally, for a program base and its EMI variant, and a program parameter m, we want to produce the same execution result, i.e.
[Base(m)]＝[Mutant 1 (m)]＝[Mutant ... (m)]＝[Mutant n (m)]
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art should be considered to be within the technical scope of the present invention, and the technical solutions and the inventive concepts thereof according to the present invention should be equivalent or changed within the scope of the present invention.
Claims (7)
1. A CPS model equivalent modulus input variation test method based on a knowledge base is characterized by comprising the following steps:
marking the step-by-step areas of all kinds of modules in the test case by using Simulink software;
establishing a deep learning model, wherein the deep learning model comprises an encoder, a decoder and a sampling mechanism, taking a correctly-operated test case as a seed model, inputting the preprocessed and coded seed model into the deep learning model, and training by using the back propagation characteristic of the deep learning model;
when the test case is mutated, the method comprises basic mutation and mutation operation, wherein when the basic mutation is: the data types of all modules of the test case and the potential conflict problem of the sample time are perfected, and then the generation operation is carried out; acquiring nested zombie areas and non-zombie areas of the test cases, randomly selecting a plurality of module sections of the non-zombie areas, inputting the module sections into a trained deep learning model, continuously writing the module sections to generate a Simulink test case capable of running correctly, and independently storing the generated test case in a knowledge base, wherein the number of the test cases in the knowledge base is increased along with the increase of variation times; during mutation operation: traversing the nested zombie areas, deleting partial modules in the nested zombie areas divided by taking the subsystem as a unit, adding test cases randomly selected from a knowledge base in the subsystem, and traversing all the areas to obtain variant test cases;
testing was performed on Simulink software: setting quantitative variation times, carrying out differential test on the variant test case and the original Simulink test case, compiling and running the test case, obtaining the input and output values of each module in the test case, comparing all the input and output values of the variant test case and the original Simulink test case, if the difference is generated, recording the difference information, otherwise, completing the test process.
2. The method of claim 1, wherein: when the step-by-step areas are obtained, the execution coverage rates of all modules are firstly obtained, so that the variant modules meeting the conditions are judged and defined as zombie blocks, the areas where the zombie blocks are located are used as variant areas input by equivalent modules, the hierarchical structures of the modules are classified, the zombie blocks located at the top layer are removed to obtain middle area zombie blocks, and the middle areas are used as nested zombie areas.
3. The method of claim 1, wherein: and (3) establishing a deep learning model by adopting a python language and a Tensorflow framework.
4. The method of claim 1, wherein: when the seed model is preprocessed: acquiring a text sequence of a test case, deleting default configuration information, default parameter sequence, blank space, annotation and position information in the text sequence, and then interleaving module and connecting line information; when the seed model is coded: and mapping words of common keywords and parameter names to numbers, and mapping each character of the rest text sequences to the numbers respectively to finally obtain the coded text sequences represented by the numbers.
5. The method of claim 1, wherein: when the potential conflict problem of the data type and the sample time of the module is perfected: and adding a type conversion module in front of each module to annotate the type of each module, and marking the sample time parameter of the signal generation module.
6. The method of claim 1, wherein: when the module segment is continuously written to generate the Simulink test case which can correctly run: and acquiring the partial module, writing the partial module into a new file, acquiring a module segment in an mdl format through the file storage characteristic of Simulink software, reading a text sequence of the mdl file, inputting the text sequence into a deep learning model for completing training, and generating a new test case by using a sampling mechanism of the deep learning model and independently storing the new test case in a knowledge base.
7. The method of claim 1, wherein: when carrying out mutation operation: traversing each nested zombie region, randomly selecting zombie blocks in the region, judging whether the zombie blocks are IF condition blocks or subsystem modules, not deleting, randomly deleting other nested zombie region blocks, storing the front and back connection information of the deleted modules, randomly loading the model in a knowledge base, adding all modules and connecting lines of the whole model in the region, connecting the head and tail modules of the model to the front and back connection positions of the deleted modules, repeating the operations until all nested zombie modules are processed, and obtaining the variant test case.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210451298.2A CN114896144B (en) | 2022-04-26 | 2022-04-26 | CPS model equivalent mode input variation test method based on knowledge base |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210451298.2A CN114896144B (en) | 2022-04-26 | 2022-04-26 | CPS model equivalent mode input variation test method based on knowledge base |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114896144A true CN114896144A (en) | 2022-08-12 |
| CN114896144B CN114896144B (en) | 2024-06-14 |
Family
ID=82720646
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210451298.2A Active CN114896144B (en) | 2022-04-26 | 2022-04-26 | CPS model equivalent mode input variation test method based on knowledge base |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114896144B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116166538A (en) * | 2022-12-28 | 2023-05-26 | 山东大学 | Cross-version prediction variation testing method and system |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060036656A1 (en) * | 2004-08-12 | 2006-02-16 | National Instruments Corporation | Automatic versioning and data mutation of user-defined data types |
| CN104391791A (en) * | 2014-11-21 | 2015-03-04 | 北京经纬恒润科技有限公司 | Method and device for testing embedded control algorithm |
| US20180349599A1 (en) * | 2017-06-06 | 2018-12-06 | Microsoft Technology Licensing, Llc | Enriching netflow data with passive dns data for botnet detection |
| CN111539099A (en) * | 2020-04-17 | 2020-08-14 | 北京航空航天大学 | Simulink model verification method based on program variation |
| CN112416806A (en) * | 2020-12-09 | 2021-02-26 | 西北大学 | JS engine fuzzy test method based on standard document analysis |
| CN113672508A (en) * | 2021-08-17 | 2021-11-19 | 大连海事大学 | Simulink test method based on risk strategy and diversity strategy |
-
2022
- 2022-04-26 CN CN202210451298.2A patent/CN114896144B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060036656A1 (en) * | 2004-08-12 | 2006-02-16 | National Instruments Corporation | Automatic versioning and data mutation of user-defined data types |
| CN104391791A (en) * | 2014-11-21 | 2015-03-04 | 北京经纬恒润科技有限公司 | Method and device for testing embedded control algorithm |
| US20180349599A1 (en) * | 2017-06-06 | 2018-12-06 | Microsoft Technology Licensing, Llc | Enriching netflow data with passive dns data for botnet detection |
| CN111539099A (en) * | 2020-04-17 | 2020-08-14 | 北京航空航天大学 | Simulink model verification method based on program variation |
| CN112416806A (en) * | 2020-12-09 | 2021-02-26 | 西北大学 | JS engine fuzzy test method based on standard document analysis |
| CN113672508A (en) * | 2021-08-17 | 2021-11-19 | 大连海事大学 | Simulink test method based on risk strategy and diversity strategy |
Non-Patent Citations (2)
| Title |
|---|
| CHOWDHURY, SHAFIUL AZAM: "Automated Testing of a Commercial Cyber-Physical System Development Tool Chain", 31 December 2019 (2019-12-31) * |
| 陈丽琼;刘微;何心;: "基于LSTM神经网络的CPS软件可靠性预测", 计算机工程与设计, no. 05, 15 May 2019 (2019-05-15) * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116166538A (en) * | 2022-12-28 | 2023-05-26 | 山东大学 | Cross-version prediction variation testing method and system |
| CN116166538B (en) * | 2022-12-28 | 2023-08-22 | 山东大学 | Cross-version prediction variation testing method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114896144B (en) | 2024-06-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Tang et al. | Convolutional neural network‐based data anomaly detection method using multiple information for structural health monitoring | |
| CN104598611A (en) | Method and system for sequencing search entries | |
| CN111949535B (en) | Software defect prediction device and method based on open source community knowledge | |
| CN111694917A (en) | Vehicle abnormal track detection and model training method and device | |
| CN112001496A (en) | Neural network structure searching method and system, electronic device and storage medium | |
| CN116204770B (en) | Training method and device for detecting abnormality of bridge health monitoring data | |
| CN114896144B (en) | CPS model equivalent mode input variation test method based on knowledge base | |
| CN110888798A (en) | Software defect prediction method based on graph convolution neural network | |
| CN110580213A (en) | Database anomaly detection method based on cyclic marking time point process | |
| CN117236677A (en) | RPA process mining method and device based on event extraction | |
| CN113902129A (en) | Multi-mode unified intelligent learning diagnosis modeling method, system, medium and terminal | |
| CN116205482A (en) | Important personnel risk level assessment method and related equipment | |
| CN113591971A (en) | User individual behavior prediction method based on DPI time series word embedded vector | |
| CN115935372A (en) | Vulnerability detection method based on graph embedding and bidirectional gated graph neural network | |
| Yang et al. | A two‐stage data cleansing method for bridge global positioning system monitoring data based on bi‐direction long and short term memory anomaly identification and conditional generative adversarial networks data repair | |
| CN113609488B (en) | Vulnerability detection method and system based on self-supervised learning and multichannel hypergraph neural network | |
| CN113076235A (en) | Time sequence abnormity detection method based on state fusion | |
| CN117272232A (en) | Tunnel monitoring method and device for data fusion, computer equipment and storage medium | |
| CN116777695A (en) | Time sequence convolution knowledge tracking method for fusion project reaction | |
| CN112735604B (en) | Novel coronavirus classification method based on deep learning algorithm | |
| CN114860952A (en) | Graph topology learning method and system based on data statistics and knowledge guidance | |
| Zhu et al. | How Robust Is a Large Pre-trained Language Model for Code Generationƒ A Case on Attacking GPT2 | |
| CN112766410A (en) | Rotary kiln firing state identification method based on graph neural network feature fusion | |
| CN114925613B (en) | Equivalent evaluation method for software reliability in space radiation environment based on deep learning | |
| CN117290856B (en) | Intelligent test management system based on software automation test technology |
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 |