59 results about "Domain testing" patented technology

The invention provides a mechanical equipment intelligent fault diagnosis method based on a partial migration convolutional network. The method comprises: collecting operation data of mechanical equipment under different operation working conditions; constituting data sets, taking part of data in the data set X as a source domain training sample set and a target domain test sample set; performingdata standardization on each piece of sample data; training two one-dimensional convolutional neural network models with the same structure and different initialization parameters by using the sourcedomain training sample set, and correcting the two trained convolutional neural network models based on the target domain test sample set to obtain a convolutional neural network mechanical equipmentfault diagnosis model; and performing fault diagnosis on the mechanical equipment based on the real-time operation data by using the fault diagnosis model to output a fault type. The method can be effectively used in more real mechanical fault diagnosis, that is to say, the label-free property of the target domain is considered, so that the trained diagnosis model can better diagnose faults of mechanical equipment.

The invention belongs to the technical field of ferroelectric or piezoelectric material electric domain testing, and relates to a double-frequency multichannel synchronization detection method for achieving electric domain imaging through an inverse piezoelectric effect. The method employs two phase locking amplifiers to achieve the real-time synchronous detection of out-plane and in-plane piezoelectric signals of a to-be-tested sample. Reference signals of the two phase locking amplifiers are provided by two independent frequency sources, and the frequency sources also provide AC signals which are the same as the reference signals, wherein the AC signals serve as excitation signals. The excitation signals provided by the two frequency sources are superposed through an adder and then serve as AC excitation signals located between a conductive probe and the to-be-tested sample. The method can detect the amplitude and phase signals of the out-plane and in-plane piezoelectric vibration of the to-be-tested sample through one-time scanning, achieves multichannel synchronous output in a real-time mode, and greatly improves the detection efficiency. Meanwhile, the method can improve the resolution of electric domain imaging, and effectively improves the detection accuracy.

The invention discloses an SAR image target detection method based on an unsupervised domain adaptive CNN, and provides the SAR image target detection method based on the unsupervised domain adaptiveCNN for SAR image target detection mainly aiming at the defects in the prior art. The method comprises the following steps: (1) generating a source domain data set; (2) generating a target domain training set and a target domain test set; (3) constructing a multi-layer feature domain adaptive network; (4) training a cyclic consistency generative adversarial network; (5) training a multi-layer feature domain adaptive network; (6) training the Faster R-CNN by using an iterative pseudo-marking method; and (7) performing position detection on the test SAR image in the target domain test set. By means of knowledge of marked source domain data, the method has the advantages that the accuracy is high, and marked SAR images do not need to be used for training target detection in a target domain.

The invention relates to a lithium ion battery SOC estimation method based on deep-transfer learning. The method comprises the steps of obtaining a source domain training set, a target domain trainingset and a test set; constructing a lithium ion battery SOC estimation source domain model based on deep learning, training the lithium ion battery SOC estimation source domain model by using the source domain training set, and storing model training data parameters; constructing a lithium ion battery SOC estimation target domain model based on deep learning, transferring lithium ion battery SOC estimation source domain model training data parameters to the lithium ion battery SOC estimation target domain model by adopting a transfer learning method, and sharing model weight parameters to perform initialization setting; and importing the lithium ion battery target domain training set into the lithium ion battery SOC estimation target domain model to perform fine adjustment training processing, and further importing the target domain training set into a target domain test set to predict the SOC value of the lithium ion battery. According to the method, the training time of the SOC estimation model of the lithium ion battery is shortened, and a large amount of time and capital investment consumed in the experimental data collection process are reduced.

The invention discloses an unsupervised domain adaptive fault diagnosis method, which comprises the steps of obtaining bearing source vibration data, and dividing the bearing source vibration data into a training sample and a test sample; constructing a CACD-1 DCNN model, training the model, determining model parameters, and performing the fault diagnosis; the source vibration data comprises target data without labels and source domain data with labels; the acquired bearing source vibration data is acquired through a sensor; an objective function of the model provided by the invention comprises cross entropy classification loss of a source domain, based on center discrimination loss and related alignment loss based on source domain and target domain features, the last two losses are executed in the last full connection layer, representation learned in a source domain can be applied to a target domain after model training, it can be guaranteed that the extracted domain invariant features have better intra-class compactness and inter-class separability, and meanwhile the extracted features can effectively improve the performance of cross-domain testing.

The invention discloses an unsupervised cross-domain action recognition method (CAFCCN) based on channel fusion and classifier confrontation. Efficient action recognition of a target domain test set based on a source domain labeled data set and a target domain unlabeled training set is achieved. The method comprises the following specific steps of: (1) selecting an action recognition model; (2) optimizing a double-flow deep network structure; (3) constructing an objective function based on the double-flow network; (4) building an unsupervised cross-domain action recognition model based on thedouble-flow network; and (5) constructing a data set. The method has the advantages that unlabeled data sets of other training sets can be subjected to efficient action recognition on the basis of theknown data set, and the problem of unlabeled data of the training set of the target data set can be effectively solved. By applying the confrontation method, confusion of categories and domains can be achieved at the same time, domain-level and class-level invariant features are obtained, the convergence speed of the method is high, and efficient recognition of actions can be achieved.

The invention discloses a method for predicting the residual life of a rotating machinery under multiple working conditions based on a dynamic domain adaptation network. The method comprises the following steps: 1, generating a source domain sample set and a target domain sample set; 2, preprocessing vibration signals in the source domain sample set and the target domain sample set; 3, generating a target domain training set and a target domain test set; 4, selecting a source domain training set by adopting a reverse verification technology; 5, constructing a dynamic domain adaptive neural network which structurally comprises a feature extractor, a prediction learning module, a marginal distribution adaptive module and a conditional distribution adaptive module; 6, training the dynamic domain adaptive neural network to obtain a trained dynamic domain adaptive neural network model; and 7, predicting the residual life of a target domain test set by using the model. According to the method, the generalization ability and the prediction precision of the residual life prediction model are improved under the condition of multiple working conditions.

The invention discloses a PTP clock synchronization precision test method for a linux system, and belongs to the field of network communication and testing. A test scene comprises a master clock device and four slave clock devices in the same PTP domain, the master clock device and the slave clock devices perform network communication through a common switch, and PTP protocol interface communication is adopted. PTP test software is operated on the master clock device and the slave clock devices respectively to perform PTP clock synchronization, the master clock device sends a synchronization message to determine a synchronization relationship between the master clock device and the slave clock devices, then the master clock device sends a timestamp following message, the slave clock devices record a time difference between the master clock device and the slave clock devices, and the slave clock devices record a synchronization error of set time. Whether precision requirements are met or not is determined. The test method is mainly based on a synchronization precision test on software, the slave clock devices do not support hardware synchronization, there is no requirement on hardware, test scene building cost is low, and the test method is easy to implement.

The invention provides a deep channel whole-domain testing method. The method includes the following steps of firstly, calculating the mechanical characteristic and selecting the loading scheme according to the surrounding rock pressure which an experimental lining segment ring should bear at the theoretical burying depth and the dead weight of the lining segment ring; secondly, preparing a longitudinal loading counterforce frame, an outer side radial loading counterforce frame and an inner side radial loading counterforce frame, and preparing the experimental lining segment ring; thirdly, installing oil cylinders at all corresponding measuring points on the longitudinal loading counterforce frame, the outer side radial loading counterforce frame and the inner side radial loading counterforce frame based on measuring points selected in the first step, dividing the outer side oil cylinders and the inner side oil cylinder into a plurality of load groups according to the loading scheme, and driving the oil cylinders of each load group by an ejecting hydraulic device; fourthly, driving one or more of the longitudinal oil cylinders, the outer side oil cylinders and the inner side oil cylinders according to the loading scheme, and conducting stepped loading on the experimental lining segment ring. The method is high in precision and can effectively restrain the fine adjustment of a PID controller.

The invention belongs to the field of artificial intelligence machine learning, and discloses a domain generalization and domain self-adaption method based on data expansion consistency, which comprises the following steps: S1, according to task requirements, designing a prediction model p theta (y | x) based on a deep neural network, theta being a model parameter, and model output being conditional probability distribution of marking y under the condition of a given sample x; S2, constructing a data expander according to task characteristics, converting the sample, and keeping the core content of the sample unchanged, so as to keep the real mark of the converted sample unchanged; S3, constructing a multi-task loss function consisting of supervised loss and data expansion consistency lossby utilizing the original training sample and the expanded sample, and training to obtain p theta * (y | x); and S4, applying the trained model p theta * (y | x) to a target field test sample for prediction. The domain generalization and domain adaptive learning method is simple, universal and good in performance, and the technical problem of domain offset can be better solved.

The invention relates to a WIFI radio frequency test system and method, and belongs to the field of automatic test technology, the test system comprises a control server, a shielding box, a parameter test device, an attenuator and a high frequency probe; the parameter test device is used for measuring an emission index of an emission signal of a tested product and outputting a second signal; the control server is provided with factory test software; the input end of the attenuator is connected with the high-frequency probe through a radio frequency cable; the high-frequency probe is arranged in the shielding box and is used for being connected with an RF head of the tested product; the shielding box is used for bearing the tested product; and based on the test system, the invention provides the WIFI radio frequency test method. Compared with the prior art, the method and the device have the effect of improving the test efficiency under the condition of ensuring relatively low cost.

The invention discloses a vulnerability type guiding fuzz testing method and system based on byte sensitive energy distribution, and belongs to the technical field of software fuzz testing. The test method comprises the following steps: marking line number information of different types of vulnerabilities through a static analysis tool; compiling an instrumentation program to realize statistics of different types of vulnerability feature information during operation; constructing and maintaining seed queues of a plurality of specific vulnerability types based on the vulnerability types, and customizing different energy distribution modes for each seed queue according to the vulnerability types; energy distribution is further refined to a byte level, and weights are distributed for variation byte positions according to potential performance scores of seeds on specific types of vulnerabilities before and after variation. By means of the byte-level energy distribution algorithm matched with the vulnerability characteristics, the vulnerability mining efficiency of the fuzzy test tool on different types of vulnerabilities is greatly improved.

The invention provides a test result classification model training method which is applied to the financial field, the artificial intelligence field or other fields. The test result classification model training method comprises the steps that a plurality of log texts are obtained, and each log text comprises a case of automatic test execution failure; a training sample data set is generated according to the multiple log texts, and the training sample data set comprises a word vector matrix and label information of the word vector matrix; and the to-be-trained test result classification model is trained by using the training sample data set to obtain a test result classification model for classifying the automatic test results. The invention further provides a test result classification method, a test result classification model training device, a test result classification device, electronic equipment, a computer readable storage medium and a computer program product.

A method and system for improving the software programming of a robot system, comprising monitoring of plurality of human user-robot interactive pairs' (HURIP) interactions. System comprises each of said plurality of HURIPs as using ‘front-end’ semi-autonomous robot component linked by wireless two-way communications to a ‘back-end’ cloud-based computerized component. Monitoring comprises review of robot sensor-gathered data and data from camera and audio data from homes of users during user-robot interactions. Analysis of said monitoring by authorized observers such as psychologist, parent, teacher, system administrator, software programmer(s), enables identification of areas for software improvement. Improved software is tested, wherein testing comprises at least similar monitoring of HURIPs, and wherein said testing comprises social robots comprising said updated software. Cycles of such monitoring of HURIP interactions, analyzing data derived from said monitoring, focus for improvement derived thereof and followed-up in coding updates, testing of updates comprising use within monitored HURIP interactions, such cycles are applied in herein disclosed method to manufacture progressively improved code for and uses of social robot system.

The invention provides a test method, and belongs to the technical field of computers. The test method comprises the steps of obtaining information of a to-be-tested system call interface; generatinga test template for testing the system call interface according to the information of the system call interface and a set generation rule, the test template which comprises a test instruction for performing a call test on the system call interface; and sending the test template to at least one to-be-tested device through the test server, and indicating each to-be-tested device to perform a call test on a local system call interface according to the test instruction in the test template. Therefore, on one hand, test instructions of all to-be-tested system call interfaces can be generated systematically; and on the other hand, for a plurality of to-be-tested devices needing to be tested, the same test template can be used for testing a plurality of local system call interfaces, batch testingof the system call interfaces of the to-be-tested devices is achieved, and therefore the testing efficiency of the system call interfaces is effectively improved.

The invention discloses a multi-source distillation-migration mechanical fault intelligent diagnosis method based on high-order moment matching, and the method comprises the steps: building a multi-source data set through the operation data collected from a plurality of mechanical devices, carrying out the preprocessing, and dividing the multi-source data set into a source domain data set, a target domain training data set, and a target domain test data set; constructing a multi-source distillation-transfer learning network model based on high-order moment matching, and performing high-order moment matching, maximum classifier difference and multi-source distillation training by using the source domain data set and the target domain training data set; and taking the target domain test data set as test input, and synthesizing outputs of the plurality of classifiers by using an adaptive weighting strategy to complete cross-domain fault diagnosis. According to the method, features of a source domain and a target domain are aligned at domain and category levels by utilizing multi-source data, the classification capability of the model on target samples is improved through multi-source distillation, and adaptive weighting is provided to integrate diagnosis results, so that the problem that the performance of a traditional method is reduced in cross-domain diagnosis is solved, and the performance of a deep model is greatly improved.

The invention discloses a new fault diagnosis method for a rotating machine based on a deep confrontation convolutional neural network. The method comprises the following steps: constructing a source domain sample data set and a target domain sample data set; constructing a deep adversarial convolutional neural network for identifying known faults and new faults, wherein the deep adversarial convolutional neural network comprises a feature extractor G, a label classifier CF, a domain discriminator D and a non-adversarial domain discriminator; in the training stage, data of a source domain and a target domain are mapped into a high-dimensional feature space through a feature extraction module, and data feature distribution is obtained; a weighted discrimination mechanism is designed, the similarity between target domain sample data and source domain data is evaluated, and the mobility of the data is discriminated; and inputting target domain test data into the trained network for testing, judging whether the data belongs to a new fault category or not through a calculated weight value, and outputting a final classification diagnosis result. Through weighted adversarial training and target domain test sample weight threshold selection, the constructed network is enabled to be suitable for known fault and new fault detection under variable working conditions.

One technical scheme of the invention is to provide a cooperative robot reachable domain test system based on LABVIEW. Another technical scheme of the invention is to provide a cooperative robot reachable domain test method based on the LABVIEW. The cooperative robot reachable domain test method based on the LABVIEW provided by the invention comprises the steps of based on a semi-physical simulation technology and a robot control technology, using a robot reverse kinematics function for calculating whether a space coordinate point in a range is reachable or not, and verifying the reachable domain of a robot according to the actual running state and the position arrival condition of the robot; besides, through a mode of forming a coordinate point set, realizing automatic traversal of coordinate points and recording of reachable points by using the semi-physical simulation technology, and detecting the reachable problem of the robot between two points and the reachable problem of a single point within a certain range through directly calling the robot reverse kinematics function, so that the test efficiency is improved; and finally, calculating the space reachable coverage rate of the robot according to the reachable point recording of the reachable domain test.

The invention provides a performance testing method and device, equipment and a storage medium, and relates to the field of software testing. After the test equipment receives a service request sent by the server, executing a target service, and then obtaining corresponding service data when the target service is executed; and determining the performance of the server based on the business data and a preset business data base value. According to the method, the business data base value can be preset, and after the business data is acquired, the test equipment can automatically analyze the acquired business data based on the preset business data base value so as to determine the performance of the server, so that a large amount of human resources can be saved, and the accuracy of the test result can be ensured.

The invention provides a test case automatic generation method and device, electronic equipment, a medium and a computer program product. The method and device can be applied to the technical field of artificial intelligence. The test case automatic generation method comprises the steps that m judgment conditions of a tested program statement are obtained, and m is an integer larger than or equal to 1; determining a test case of each judgment condition according to the m judgment conditions; according to the test cases, an effective test case group of the tested program statement is determined, and the effective test case group is a set of test cases capable of covering and testing the m judgment conditions; and calling a test case optimization model, and determining a case test value in the effective test case group.