34results about How to "Reduce false alarm probability" patented technology

A method for notification in communication systems is disclosed, in which a sequence of K notification indicators is assigned to a notification identifier like for example a MBMS service identifier or service group identifier of UMTS. The K notification indicators are transmitted one per frame in a cyclic manner. To obtain the notification indicator identifiers indexing the notification indicators in each frame, the plurality of notification identifiers is arranged in a K-dimensional space. Each of the K coordinates of this space can assume Nni values, where Nni is the number of notification indicators signalled per frame. The i coordinate Xi,n of the point to where the n notification sequence is mapped in the K-dimensional space, forms the identifier for the notification indicator belonging to the n notification sequence in the frame with the frame number satisfying the equation i = M mod K. <IMAGE>

The invention discloses a wireless acoustic emission detector of a wind power generator unit. The wireless acoustic emission detector comprises acoustic emission receivers used for collecting and amplifying acoustic emission signals of bolt breakage of the wind power generator unit and extracting characteristic parameters and further comprises a host computer used for analyzing and recording the extracted characteristic parameters of the acoustic emission signals, carrying out logical decision and carrying out display alarm. A communication module used for transmitting data between the acoustic emission receivers and the host computer is arranged between the acoustic emission receivers and the host computer. According to the wireless acoustic emission detector of the wind power generator unit, traditional acoustic emission sensors, preamplifier circuits and preprocessing circuits are integrated to form the integral acoustic emission receivers, the designing scheme of adopting wireless transmission between the acoustic emission receivers and the host computer is adopted, and therefore the shortcoming of arranging wires inside the wind power generator unit is overcome; and the redundancy design of arranging two acoustic emission receivers at each detection point is adopted, the results generated by two acoustic emission receivers are jointly judged, and therefore the occurrence probability of missed alarm and false alarm can be reduced, and the monitoring result can be more accurate.

The invention discloses a face recognition campus area early warning method and system for a K12 education stage, and the method comprises the steps: defining an area needing to be managed and controlled, and setting a safety and early warning range; judging whether a person appearing in the video is an internal person or not, analyzing the advancing direction of a human body appearing in the video, and judging whether the person has a trend of entering a dangerous area or not; and discriminating the personnel entering the management and control area, and reminding security personnel of payingattention to the abnormal state of the management and control area. According to the invention, a region identification mode is adopted; based on face recognition and behavior analysis, people entering a management and control area are doubly identified, human bodies and objects can be clearly identified, adults and children and people outside and inside a school are further distinguished, the false alarm rate in an existing dangerous area early warning technology is effectively reduced, an alarm is given in a targeted mode, and the early warning accuracy is improved.

The invention provides an early warning method and device for equipment performance and monitoring equipment. The early warning method comprises the steps: determining S performance periods of to-be-monitored equipment in a sampling time period according to the change trend of the performance data, collected in the sampling time period, of the to-be-monitored equipment along with the time; dividing any one performance period Ti in the S performance periods into m sub-time intervals, and determining a first fluctuation interval corresponding to the jth sub-time interval according to the performance data of the to-be-monitored equipment in the jth sub-time interval in each performance period; and determining a second fluctuation interval corresponding to the jth sub-time interval according to the similarity between the Tijth sub-time interval and each Tkjth sub-time interval. According to the early warning method, the current performance data of the to-be-monitored equipment is collected, and whether the performance of the equipment is early warned or not is determined according to the current performance data, the first fluctuation interval and the second fluctuation interval, so that the early warning missing probability is reduced while the early warning range is expanded, and the early warning accuracy is improved.

The present invention is characterized in a method comprises a first step of arranging data positions as cell data forms by using distances as indexes by a register after inputting Range Data of a radar signal; a second step of scanning the register by a Sliding Window, assigning at least more than one Test Cell, and assigning at least more than one Reference Cell; a third step of respectively comparing data values of the Test Cells and data values of the Reference Cells by a comparator; and a fourth step of processing the Test Cells having the data values smaller than that of the Reference Cells as noises, according comparative results of the comparator, by determining the Test Cells having the data values larger than that of the Reference Cells as target data.

The invention relates to a method for fault diagnosis and intelligent early warning of monitoring system equipment, comprising the following steps: step 1, establishing independent alarm analysis logic and recovery analysis logic based on each alarm fault model; step 2, controlling the alarm logic and The recovery logic is mutually exclusive; Step 3: Delay processing of the alarm event and the recovery event; Step 4: Establish a cascade relationship for the alarm events generated after the delay processing: according to the relationship between different alarm events, establish the alarm level Cascading groups, with the occurrence of alarm events as trigger conditions, automatically judge whether the same cascading group alarms exist at the same time within a certain period of time, and generate correlation information between alarm events. Compared with the prior art, the present invention has the advantages of reducing the probability of false alarms, reducing the number of invalid alarms, reducing storage space overhead, reducing system CPU consumption, and reducing the workload of maintenance personnel.

The invention provides a business resources monitoring method and device, computing equipment and a storage medium. The method comprises the following steps: updating a monitoring threshold by using a preset updating logic; wherein the change of the monitoring threshold value is in positive correlation with the change of the business resources; and in each monitoring period, using the updated monitoring threshold value to monitor the business resources. According to the scheme, the false alarm probability can be reduced.

The invention relates to a fault identification and positioning method and system of an intelligent photovoltaic power generation system. The fault identification and positioning method comprises the following steps: S1, obtaining an association coefficient of a photovoltaic power generation unit relative to other photovoltaic power generation units; S2, establishing a fault determination basis of the photovoltaic power generation unit; and S3, carrying out fault identification and positioning on a faulted photovoltaic power generation unit. According to the invention, on the basis of discrimination of distances among all units and real-time combined generation efficiencies of all the units of the photovoltaic power generation system, data analyses are carried out to realize accurate identification and positioning of faults of the intelligent photovoltaic power generation system, thereby substantially reducing the false fault alarm probability and improving fault identification accuracy and fault positioning precision of the intelligent photovoltaic power generation system. Meanwhile, no extra professional fault diagnosis device is needed for the photovoltaic power station and thus costs are lowered. The method and system have broad application prospects.

The embodiments of the invention disclose an application layer slow-speed attack detection method and a correlation apparatus. The application layer slow-speed attack detection method comprises the following steps: obtaining a flow entering a server; based on the obtained flow entering the server, counting the quantity of transmission layer empty chains initiated to the server within preset duration; and if the counted quantity of the transmission layer empty chains is greater than or equal to a first threshold and the quantity of application layer data packets carrying a destination internetworking protocol address i and a public network gateway interface identification j in the flow entering the server within k periods accords with a preset threshold condition, performing application layer slow-speed attack alarm on the destination internetworking protocol address i and the public network gateway interface identification j, wherein the destination internetworking protocol address i is one of internetworking protocol addresses of the server. According to the technical scheme provided by the embodiments of the invention, the accuracy of the application layer slow-speed attack alarm is improved.

The application provides an early warning method, device, and monitoring equipment for equipment performance. The method includes: determining the S performance period; and any performance period T in the S performance period i Divide it into m sub-time intervals, determine the first fluctuation interval corresponding to the j-th sub-time interval according to the performance data of the equipment to be monitored in the j-th sub-time interval in each performance cycle; then according to the T ij sub-time intervals and each Tth kj The similarity between the sub-time intervals determines the second fluctuation interval corresponding to the j-th sub-time interval; collects the current performance data of the equipment to be monitored, and determines whether the equipment is correct according to the current performance data, the first fluctuation interval and the second fluctuation interval. The performance of the early warning, while expanding the scope of the early warning, reduce the probability of missed warning, thereby improving the accuracy of the early warning.