31 results about "Nonnormal data" patented technology

The invention relates to multilayer honey-net data transmission method and system, and the gateway of honey-net receives the data flow of external network; the gateway of honey-net detects the network intrusion for the received data flow; the normal data flow can pass, which is sent to the target host computer; and the informal data flow is divide into high, middle and low according to dangerous level; and the data flow with high level is sent to physical honey-tank system, the data flow with middle level is sent to virtual machine honey-tank system, and the data floe with low level is sent to virtual honey-tank system. The invention can utilize the advantages of low interactive honey-tank system and high interactive honey-tank system, save system source, and improve the covering area of honey-tank system and the ability of obtaining the movable information of network attack and capturing the malice code; it can defeat the anti-honey tank technique. And it can be used in security field of computer network.

The invention relates to the technical field of software testing, in particular to a data source configuration testing method, system, a computer device and a storage medium. The data source configuration testing method comprise that following steps: obtaining a first normal data set accord with a normal flow from the data source, starting a data generation tool to process the first normal data set to generate an abnormal data set with abnormal data; obtaining a second normal data set conforming to a normal flow from the data source, extracting a part of data in the abnormal data set and encapsulating the part of data in the second normal data set into a layer of verification use case framework; the interceptor is called to test the abnormal data, and the test results are brought into theon-going service process, and the accuracy of the test results is verified by observing the results of the service process running. By grouping and integrating the data, the invention can automatically complete the coverage of the abnormal scene, and does not need manual intervention in the regression test.

The invention provides an optical fiber vibration-based railway perimeter protection method. The method comprises the following steps that: a railway perimeter protection system is set, wherein the protection system includes defense area induction modules, defense area data acquisition units, an information processing host and a railway administration monitoring platform; induction head end modules and induction tail end modules are distributed at defense areas of railway perimeter along a railway, each defense area induction module forms a defense area; each defense area data acquisition unit acquires intrusion vibration signals of the plurality of defense area induction modules; the information processing host receives the intrusion vibration signals of the defense area data acquisition units and identifies the intrusion vibration signals, and distinguishes abnormal data and transmits the abnormal data in a packeted manner; and the railway administration monitoring platform is used for receiving the alarm data of the information processing host and notifying responsible personnel of the alarm data.

The invention relates to an intelligent remote warehouse data monitoring alarm system and relates to an alarm system. The system comprises a sensor data acquisition device, an intelligent monitoring service platform and a mobile terminal, wherein the sensor data acquisition device comprises a system host and a remote controller, the host has a rectangular casing, a front surface of the host is provided with a liquid crystal display screen, a left lower corner of the front surface is provided with an infrared receiver, an upper portion of a right side surface of the host is a temperature, humidity and smog sensor acquisition data window, a lower portion of the right side surface of the host is a buzzer, an upper left side of the host is covered by a solar energy photovoltaic panel, an upper right side of the host is provided with a flash indication lamp close to the display screen, and an antenna is arranged at the position of the upper right side of the host away from the display screen. The system is advantaged in that temperature, humidity and smog data of a warehouse is acquired through sensors, the data is sent through a ZigBee to an intelligent monitoring platform of a monitoring center in a wireless mode, and non-normal data or advice for a warehouse with possible problems and alarm are provided by the intelligent monitoring platform according to data analysis and data mining of the big data.

The invention relates to a rapid navigation channel model modeling method for 3D printing. The method comprises steps that initial terrain data of the engineering river section including the terrain elevation is collected; a model making scope is determined; abnormal data of the initial terrain data is excluded; the model both-side flat top elevation is determined; the model both-side flat top elevation is utilized to replace the shore elevation; the model bottom elevation is determined; model boundary control point data is integrated; discrete processing is carried out; a navigation channel model is generated, the model is cut into a plurality of channel model blocks, and the model blocks are outputted and printed. The method is advantaged in that a digital model which can be identified by a 3D printer can be rapidly established for any complex river terrains, and the river terrain model can be guaranteed to have the relatively high similarity to the prototype.

The invention discloses a method for extracting data of BlueSky file system monitoring equipment rapidly. The method includes the steps of S1, loading disk data; S2, detecting whether the disk data meet requirements or not; S3, analyzing disk structure; S4, analyzing normal data records; S5, analyzing abnormal data records; S6, classifying and extracting video data; S7, recombining fragmented data. The method has the advantages that normal video, deleted video and fragmented video of a BlueSky file system can be extracted completely and rapidly, and the fragmented data can be recombined, thereby helping in data recovery of enterprises and case detection of the public security organization.

The embodiment of the invention discloses a UWB (Ultra-wide Bandwidth) based alarming method and equipment. The method comprises the following steps: acquiring real-time position data of different positioning tags within a preset time period on the basis of a UWB channel, wherein the positioning tags are arranged on livestock, and the real-time position data are used for indicating movement rangesand movement frequencies of the positioning tags; according to the real-time position data, confirming types of living evaluation data for evaluating living situations of the livestock, wherein the types of the living evaluation data comprise question data, suspected question data and normal data; and when the types of the living evaluation data are abnormal data, acquiring and outputting targettag IDs (Identities) of target positioning tags corresponding to the real-time position data. By adopting the method, the positioning tags are arranged on the livestock by using a UWB positioning system, position data of the livestock are acquired and analyzed on the basis of the UWB channel, then the living situations of the livestock can be confirmed, and furthermore intelligent breeding requirements of medium/small-size breeding farmers can be met.

The invention discloses a risk early warning method for electrical equipment of a mountain transformer substation, and belongs to the technical field of electric power. According to the invention, abnormal data is corrected; meanwhile, normalization is carried out to realize dimensionless processing, and the processed data is processed by adopting an LSTM algorithm so that prediction precision canbe effectively enhanced, a manager is enabled to acquire more intuitive and clear equipment temperature change conditions, prediction can be realized, state switching can be carried out through a finite state machine, and thus risk early warning and generator tripping processing can be realized.

The invention discloses a fruit weighing error control method. The method comprises the steps of conveying by a conveying belt, weighing by a weighing platform, sensor data acquisition, data preprocessing and data comprehensive analysis. A pressure sensor, a vibration sensor and a tilt angle sensor are arranged to acquire the data, and then the corresponding data is preprocessed and comprehensively analyzed; invalid data is filtered and removed, abnormal data with a large numerical value difference is deleted, interference of individual data on a final result is reduced, the accuracy of a calculation result is further improved, then influences caused by vibration and angle inclination are comprehensively considered, the quality of fruits is automatically obtained through an algorithm, interference of multiple influence factors is avoided, accurate data support is provided for quality classification of fruits, the labor intensity of workers is reduced, and time is saved.

The invention discloses a merging unit equipment state prediction method based on time sequence analysis. The method comprises the following steps of (1), preparing data, and constructing a data set;(2) trend analysis of normal data and abnormal data; (3) analyzing a time sequence and establishing a prediction model; and (4) predicting a state of a merging unit device. According to the method, through fitting the monitoring data of an operation state of the existing merging unit device, the information and trend hidden in the data are mined, maintenance cost is reduced, and troubleshooting efficiency is improved; time sequence analysis is adopted according to historical data, so prediction in the future is realized, and an alarm can be given in time when equipment runs abnormally; the prediction method of the merging unit equipment state is provided for analyzing and mastering the equipment state information.

The invention relates to a method for detecting abnormal data of an automobile maintenance record library. The method comprises the steps of performing preprocessing and feature extraction on the data of the automobile maintenance record library; reconstructing the data by using an automatic encoder for suppressing the depth; obtaining weight vectors and bias vectors of all layers with reconstruction loss and accuracy meeting expected requirements through deep training, and constructing a discrimination standard on the basis of the weight vectors and the bias vectors to carry out normal feature labeling on normal data; and performing abnormal feature labeling on the abnormal data so as to complete the detection of the abnormal data. According to the invention, through regularization suppression of the loss function, a detection algorithm for the abnormal data of the automobile maintenance record library is established, and the accuracy of abnormal data identification and the quality of data in the library can be effectively improved.

The invention discloses a diagnosis method and device for a charging pile. The diagnosis method for the charging pile comprises the following steps: determining a safety threshold range of electric signal data of each monitoring point in the charging pile; acquiring electric signal data of each monitoring point in real time; judging whether the electric signal data of each monitoring point is within a corresponding safety threshold range or not, and if not, determining that the corresponding electric signal data is abnormal data; whether the charging pile is in a controlled charging state or not is detected, if yes, it is determined that abnormal data is abnormal data, and if not, it is determined that the abnormal data is fault data; judging whether a diagnosis rule corresponding to the abnormal data or the fault data exists or not, if yes, determining a generation reason and a solution of the abnormal data or the fault data according to the corresponding diagnosis rule, and if not, generating diagnosis-rule-free information; and sending the generation reason and the solution or the non-diagnostic rule signal to a server side. According to the technical scheme provided by the embodiment of the invention, the utilization rate of the charging pile is improved.

The invention discloses a computer networked information safety application system, and relates to the technical field of agricultural management. a data receiving module is used for receiving downloaded data for normal work; a standby module is used for detecting whether the downloaded data are normal or not; a main module is used for receiving normal data. According to the system, normal data can be received through the data receiving module, abnormal data intrusion can be prevented, data which cannot be recognized whether to be normal or not can be downloaded in advance through the standby module, whether the data is normal or not can be observed, and the safety of the data storage module can be improved through a password module and a dynamic verification module; when the data attempts to enter the data storage module, the data signals are transmitted to the information recording module and recorded as abnormal data, so that the data of the type can be prevented from entering again, and the safety of the data is further improved.

The invention belongs to the technical field of equipment remote detection, and particularly discloses a remote detection and diagnosis system for intelligent equipment which comprises a monitoring module, an information processing unit, a 5G communication module, a fault diagnosis terminal platform and a mobile terminal. The monitoring module is used for collecting data in the equipment operation process and transmitting the collected data information to the information processing unit. The information processing unit is used for analyzing and integrating data information and dividing the data information into abnormal data and normal data, and the 5G communication module is used for transmitting the abnormal data to the fault diagnosis terminal platform; the fault diagnosis terminal platform is used for generating a fault diagnosis result and a fault solution; the mobile terminal is used for transmitting the fault diagnosis result and the fault solution to a maintenance worker. According to the diagnosis system provided by the invention, an extremely complex fault detection link is omitted, the fault detection efficiency is improved, and the efficiency of a production line and the real-time safety of personnel are ensured.

The invention relates to a method and apparatus for identifying an asymmetric waveform in a strong earthquake record. The method comprises: carrying out base line initialization on acceleration data in an obtained strong earthquake record; classifying corrected acceleration data into a first array and a second array by using a base line as a benchmark; carrying out a preliminary symmetry analysis on the first array and the second array and determining whether a graph drawn by the acceleration data is a symmetrical waveform; and if so, carrying out a precise symmetry analysis on the first array and the second array and determining further whether the graph drawn by the acceleration data is a symmetrical waveform. According to the method and apparatus provided by the invention, advantages of two kinds of symmetric analysis methods are combined; and the identification efficiency can be guaranteed and the screening accuracy can be improved; abnormal data in a strong earthquake record can be rejected; and the earthquake analysis result reliability can be enhanced.

The invention relates to the field of power systems, in particular to a data cleaning method and device in soil big data analysis. The method comprises the following steps: acquiring soil data, and acquiring environment data when the soil data is acquired; performing data dispersion on the acquired soil data according to categories to obtain a plurality of dispersed data sets; on the basis of the data structure and the data size of each piece of dispersed data, constructing a dispersed data sphere; and finally, constructing a data cleaning cube, and integrating the data cleaning cube and the dispersed data spheres to obtain final cleaning data. According to the method, data cleaning is carried out by using a mode which is different from a mode that only abnormal value searching is carried out on the data in the prior art, and the normal data is marked by using a mode based on data cube construction, so that the abnormal data is corrected; and the correction model is constructed in combination with abnormity frequently caused by environmental data in the soil data, so that the accuracy of data cleaning is remarkably improved.

The invention discloses a method for predicting an occurrence risk of an acute cerebrovascular disease, and relates to the field of acute cerebrovascular diseases. The technical scheme comprises the following steps: step 1, data collection: collecting data of a large number of patients, and collecting body data of immediate family members of the patients suffering from the acute cerebrovascular disease; step 2, data processing: processing the collected data, paralleling each type of data, and removing abnormal data in the data at the same time; and step 3, carrying out grading processing on the relationship between the data after data processing, and analyzing the relevance between families at the same time. According to the method, the data of the patients is counted during data statistics, meanwhile, the data of the patients suffering from the acute cerebrovascular disease among the immediate family members of the patients is counted, and the data relevance between the patients and the immediate family members suffering from the acute cerebrovascular disease is analyzed, so that the relevance between the data can be increased, and the prediction accuracy is ensured.

The invention discloses a data risk analysis method. The method comprises the steps of obtaining basic information of an abnormal organization according to abnormal organization characteristics; constructing an abnormal database according to the basic information of the abnormal organization; and performing risk analysis on the obtained organization ticket business data according to the abnormal database to obtain a ticket business risk result. According to the invention, the abnormal database is constructed through the obtained ticket business data, and then risk analysis is carried out through the abnormal database to obtain the ticket business risk result, so that the accuracy of data risk analysis is improved. The invention further discloses a data risk analysis device, a server and acomputer readable storage medium, which have the above beneficial effects.