CN115964678B - Intelligent identification method and system based on multi-sensor data - Google Patents
Intelligent identification method and system based on multi-sensor data Download PDFInfo
- Publication number
- CN115964678B CN115964678B CN202310254438.1A CN202310254438A CN115964678B CN 115964678 B CN115964678 B CN 115964678B CN 202310254438 A CN202310254438 A CN 202310254438A CN 115964678 B CN115964678 B CN 115964678B
- Authority
- CN
- China
- Prior art keywords
- data
- characteristic
- sensor
- recognition
- layer
- 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.)
- Active
Links
Classifications
-
- 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
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Arrangements For Transmission Of Measured Signals (AREA)
Abstract
The application provides an intelligent recognition method and system based on multi-sensor data, which relate to the technical field of data analysis, and are characterized in that a sensor data sharing platform is built, information of a first recognition target is acquired, the information is input into the sensor data sharing platform, a plurality of primary sensing data sources are acquired according to a sensor network layer, the primary sensing data sources are transmitted to a feature recognition layer according to a data transmission layer, a plurality of data feature intensities are acquired, a plurality of secondary sensing data sources are determined, the secondary sensing data sources are combined and output through a data output layer, so that the technical problems that the recognition method of the multi-sensor data in the prior art is most innovated in the technical aspect, the feature intensities of the sensing data sources are ignored, the data recognition efficiency is low and the recognition accuracy is insufficient are solved, the trace is carried out, the analysis is carried out based on the root sources of the sensing data sources, the feature intensities of the sensing data sources are guaranteed, the data recognition difficulty is reduced, and the recognition efficiency and the recognition accuracy are improved.
Description
Technical Field
The application relates to the technical field of data analysis, in particular to an intelligent identification method and system based on multi-sensor data.
Background
The sensor is used as equipment for collecting and measuring data and comprises a plurality of sensing equipment types, in general, the sensor is used for collecting data based on preset frequency and sending the data to a corresponding data receiving end, and the data processing method is used for determining the development trend and the change rule of things based on the received time sequence data, so that the requirement on the identification precision of the sensing data is higher, namely the characteristic strength of the sensing data needs to be ensured.
At present, the current traditional sensing data identification method mainly carries out data identification through technical means such as modeling, data enhancement, data mining and the like, omits evaluation analysis on sensing data sources, and strengthens the identification difficulty of sensing data.
In the prior art, many recognition methods for multi-sensor data are innovations in technical aspects, and feature intensity of a sensing data source is ignored, so that data recognition is difficult, and low data recognition efficiency and insufficient recognition accuracy are caused.
Disclosure of Invention
The application provides an intelligent recognition method and system based on multi-sensor data, which are used for solving the technical problems of low data recognition efficiency and insufficient recognition accuracy caused by difficult data recognition due to the fact that the characteristic intensity of a sensing data source is ignored because of the innovation of most of the technical aspects of the recognition method of the multi-sensor data in the prior art.
In view of the above problems, the application provides an intelligent recognition method and system based on multi-sensor data.
In a first aspect, the present application provides an intelligent recognition method based on multi-sensor data, the method comprising:
the sensor data sharing platform is built, wherein the sensor data sharing platform comprises a sensor network layer, a data transmission layer, a feature identification layer and a data output layer;
acquiring information of a first identification target;
inputting the information of the first recognition target into the sensor data sharing platform, and acquiring a plurality of primary sensing data sources based on the first recognition target according to the sensor network layer of the sensor data sharing platform, wherein the plurality of primary sensing data sources are in one-to-one correspondence with a plurality of sensors;
transmitting the plurality of primary sensing data sources to the characteristic recognition layer according to the data transmission layer to recognize the data characteristic intensity, and acquiring a plurality of data characteristic intensities;
acquiring a plurality of secondary sensing data sources according to the plurality of data characteristic intensities;
and outputting the primary sensing data source and the secondary sensing data source through the data output layer.
In a second aspect, the present application provides an intelligent recognition system based on multi-sensor data, the system comprising:
the platform building module is used for building the sensor data sharing platform, wherein the sensor data sharing platform comprises a sensor network layer, a data transmission layer, a feature identification layer and a data output layer;
the information acquisition module is used for acquiring information of the first identification target;
the primary sensing data source acquisition module is used for inputting the information of the first identification target into the sensor data sharing platform, and acquiring a plurality of primary sensing data sources based on the first identification target according to the sensor network layer of the sensor data sharing platform, wherein the plurality of primary sensing data sources are in one-to-one correspondence with a plurality of sensors;
the characteristic intensity recognition module is used for transmitting the plurality of primary sensing data sources to the characteristic recognition layer to recognize the characteristic intensity of the data according to the data transmission layer, and acquiring a plurality of characteristic intensities of the data;
the secondary sensing data source acquisition module is used for acquiring a plurality of secondary sensing data sources according to the plurality of data characteristic intensities;
and the data source output module is used for outputting the primary sensing data source and the secondary sensing data source through the data output layer.
One or more technical schemes provided by the application have at least the following technical effects or advantages:
the embodiment of the application provides an intelligent recognition method based on multi-sensor data, which is used for building a sensor data sharing platform and comprises a sensor network layer, a data transmission layer, a feature recognition layer and a data output layer; acquiring information of a first identification target, inputting the information into the sensor data sharing platform, and acquiring a plurality of primary sensing data sources based on the first identification target according to the sensor network layer of the sensor data sharing platform, wherein the plurality of primary sensing data sources are in one-to-one correspondence with a plurality of sensors; transmitting the plurality of primary sensing data sources to the characteristic recognition layer according to the data transmission layer, and recognizing the data characteristic intensity to obtain a plurality of data characteristic intensities; acquiring a plurality of secondary sensing data sources according to the plurality of data characteristic intensities; the primary sensing data source and the secondary sensing data source are output through the data output layer, so that the technical innovation of most of recognition methods of multi-sensor data in the prior art is solved, the characteristic intensity of the sensing data source is ignored, data recognition is difficult, the technical problems of low data recognition efficiency and insufficient recognition accuracy are caused, the trace source is performed, the characteristic intensity of the sensing data source is guaranteed based on the analysis of the root cause of the sensing data source, the data recognition difficulty is reduced, and the recognition efficiency and accuracy are improved.
Drawings
FIG. 1 is a schematic flow chart of an intelligent recognition method based on multi-sensor data;
FIG. 2 is a schematic diagram of a process for identifying the intensity of a plurality of data features in an intelligent identification method based on multi-sensor data;
FIG. 3 is a schematic diagram of a process for acquiring multiple secondary sensing data sources in an intelligent recognition method based on multiple sensor data;
fig. 4 is a schematic structural diagram of an intelligent recognition system based on multi-sensor data.
Reference numerals illustrate: the system comprises a platform building module 11, an information acquisition module 12, a primary sensing data source acquisition module 13, a characteristic intensity identification module 14, a secondary sensing data source acquisition module 15 and a data source output module 16.
Detailed Description
The application provides an intelligent recognition method and system based on multi-sensor data, which are characterized in that a sensor data sharing platform is built, information of a first recognition target is obtained, the information is input into the sensor data sharing platform, a plurality of primary sensing data sources are obtained according to a sensor network layer, the primary sensing data sources are transmitted to a feature recognition layer according to a data transmission layer, data feature intensity recognition is carried out to obtain a plurality of data feature intensities, and a plurality of secondary sensing data sources are obtained according to the plurality of data feature intensities; the primary sensing data source and the secondary sensing data source are output through the data output layer, so that the technical problems that the identification method of multi-sensor data in the prior art is most of innovation in the technical aspect, the characteristic intensity of the sensing data source is ignored, the data identification is difficult, the data identification efficiency is low and the identification accuracy is insufficient are solved.
Example 1
As shown in fig. 1, the present application provides an intelligent recognition method based on multi-sensor data, the method is applied to a multi-sensor data processing system, the system is in communication connection with a sensor data sharing platform, the method includes:
step S100: the sensor data sharing platform is built, wherein the sensor data sharing platform comprises a sensor network layer, a data transmission layer, a feature identification layer and a data output layer;
specifically, the sensor is used as equipment for collecting and measuring data and comprises a plurality of sensing equipment types, in general, the sensor is used for collecting data based on preset frequency and sending the data to a corresponding data receiving end, and the object development trend and change rule are determined based on the received time sequence data.
Specifically, a multi-level functional layer is constructed, wherein the multi-level functional layer comprises a sensor network layer, a data transmission layer, a feature identification layer and a data output layer, the sensor network layer is used for data integration, a plurality of network venues are contained for associating multiple sensors, the data transmission layer is used for transmitting the multi-sensor data integrated by the sensor network layer to the feature identification layer, data features are obtained and output based on the data output layer, the sensor network layer, the data transmission layer, the feature identification layer and the data output layer are sequentially and hierarchically associated to form a sensor data sharing platform, and based on the sensor data sharing platform, the identification analysis of the multi-sensor data is carried out, the standardization of an analysis flow is ensured, and the identification energy efficiency is ensured.
Step S200: acquiring information of a first identification target;
step S300: inputting the information of the first recognition target into the sensor data sharing platform, and acquiring a plurality of primary sensing data sources based on the first recognition target according to the sensor network layer of the sensor data sharing platform, wherein the plurality of primary sensing data sources are in one-to-one correspondence with a plurality of sensors;
specifically, the first recognition target is a target to be subjected to sensing data acquisition and recognition, and the information of the first recognition target is acquired, namely, the basic information of the first recognition target can be recognized and judged, wherein the basic information comprises state characteristics, morphological attributes and the like. And further inputting the information of the first identification target into the sensor data sharing platform, and carrying out information identification to determine a plurality of sensors associated with the first identification target, wherein the plurality of sensors comprise a plurality of types, and the data acquisition is carried out on the first identification target based on multiple dimensions so as to ensure the completeness of the information. And integrating the acquired data of the multiple sensors based on the sensor network layer of the sensor data sharing platform, wherein the acquired data corresponding to one sensor is used as a single primary sensing data source, a plurality of primary sensing data sources of the first recognition target are acquired, and the acquisition of the plurality of primary sensing data sources provides actual data support for subsequent recognition analysis.
Step S400: transmitting the plurality of primary sensing data sources to the characteristic recognition layer according to the data transmission layer to recognize the data characteristic intensity, and acquiring a plurality of data characteristic intensities;
further, as shown in fig. 2, the plurality of primary sensing data sources are transmitted to the feature recognition layer for data feature intensity recognition, and step S400 of the present application further includes:
step S410: transmitting the primary sensing data source to the feature recognition layer, wherein the feature recognition layer is embedded with a feature intensity recognition model;
step S420: analyzing according to the characteristic intensity recognition model to obtain characteristic data sensing granularity, characteristic data chronology and characteristic data labeling quantity;
step S430: and carrying out feature intensity weight calculation by taking the feature data sensing granularity, the feature data time sequence and the feature data labeling quantity as variables, and outputting the plurality of data feature intensities.
Specifically, the plurality of primary sensing data sources are transmitted to the feature recognition layer based on the data transmission layer, and the plurality of data feature intensities are output by performing data feature intensity recognition based on the feature intensity recognition model embedded in the feature recognition layer.
Specifically, the characteristic intensity recognition model is constructed, multiple groups of sample sensing data sources are collected, characteristic recognition is respectively carried out to determine multi-dimensional characteristics, the characteristic intensity recognition model comprises multiple groups of sample characteristic data sensing granularity, multiple groups of sample characteristic data timeliness and multiple groups of sample characteristic data labeling quantity, the multiple groups of sample sensing data sources are mapped and associated with the multi-dimensional characteristics, the constructed sample data are obtained to carry out neural network training to generate a characteristic parameter extraction layer, multiple groups of preset characteristic weights are configured, the characteristic parameter weights corresponding to different recognition targets may have differences, the multiple groups of preset characteristic weights are embedded into the intensity calculation layer, the characteristic intensity recognition model is generated based on the characteristic parameter extraction layer and the intensity calculation layer, the characteristic intensity recognition model is embedded into the characteristic recognition layer, and network layer function optimization is carried out.
Transmitting the primary sensing data source to the feature recognition layer, performing recognition and matching based on the embedded feature intensity recognition model, and acquiring the feature data sensing granularity, the feature data time sequence and the feature data labeling quantity, wherein not all the complete sensing data are feature data, only a few node data possibly exist in one data sequence and belong to the feature data, and the feature data sensing granularity is the extraction node of the feature data compared with the complete sensing data source. And the characteristic data chronology is the serialization of the extraction time nodes of the characteristics, the characteristic data marking quantity is the manual marking existing after the primary sensing data source is initially acquired, and the characteristics with the manual marking are the necessary characteristics.
And further determining a preset feature weight according to the strength calculation layer, taking the feature data sensing granularity, the feature data time sequence and the feature data labeling quantity as variables, respectively carrying out weighted calculation on different feature data, taking a calculation result as data feature strength, and integrating the data feature strengths to carry out model output. By constructing the characteristic intensity recognition model to perform characteristic intensity analysis, the analysis efficiency can be effectively improved, and the accuracy and objectivity of an analysis result are ensured.
Further, the present application also includes step S440, including:
step S441: acquiring a test sample data set;
step S442: inputting the test sample data set into the data transmission layer for analysis to obtain a transmission sample data set;
step S443: comparing the transmission sample data set with the test sample data set to obtain a transmission loss coefficient;
step S444: and adjusting the characteristic intensities of the data according to the transmission loss coefficient.
Specifically, in the process of data transmission based on the data transmission layer, certain transmission loss inevitably exists, so that slight data deviation is caused, and the finally determined plurality of data characteristic intensity precision is insufficient. And acquiring the test sample data set, namely sample data which is the same in data type and data format as the plurality of primary sensing data sources. Inputting the test sample data into the data transmission layer, and taking the transmitted test sample data set as the transmission sample data set through data transmission. And carrying out data overlap correction on the transmission sample data set and the test sample data set, determining the transmission loss degree based on the deviation degree of the transmission sample data set and the test sample data set, and acquiring the transmission loss coefficient, namely the expression data for measuring the transmission loss degree, wherein the larger the transmission loss coefficient is, the higher the loss degree in the corresponding data transmission process is. And determining the adjustment direction and the adjustment scale of the plurality of data characteristic intensities based on the transmission loss coefficient, and adjusting the plurality of data characteristic intensities. The data deviation caused by data transmission can be effectively reduced, and the accuracy of the characteristic intensity of the data can be improved.
Step S500: acquiring a plurality of secondary sensing data sources according to the plurality of data characteristic intensities;
step S600: and outputting the primary sensing data source and the secondary sensing data source through the data output layer.
Specifically, the data characteristic intensities are determined, N sensor data sources which do not meet preset characteristic intensities in the data characteristic intensities are used as data sources with the need of re-acquisition, and the data sources are sent to the sensor network layer of the sensor data sharing platform. And determining preset data acquisition amounts based on the characteristic intensity differences respectively, and generating N data acquisition instructions. Based on the N data acquisition instructions, controlling the corresponding N sensors, taking the preset data acquisition amount as an acquisition standard, carrying out sensing data acquisition, and taking an acquisition result as a plurality of secondary sensing data sources.
Preferably, the sensor data sharing platform constructed in the real-time embodiment of the present application has a plurality of closed loop paths, and the operation mechanism thereof is as follows: performing layer-by-layer analysis on the plurality of primary sensing data sources integrated by the sensor network layer, namely, the characteristic identification layer, further performing characteristic intensity judgment, and outputting the plurality of primary sensing data sources directly based on the data output layer when the characteristic intensities meet the corresponding preset characteristic intensities; when the characteristic intensity does not meet the preset characteristic intensity, determining a corresponding sensing data source, feeding back and transmitting the sensing data source to the sensor network layer by the characteristic identification layer, performing secondary acquisition of sensing data, performing data fitting by combining the primary sensing data source, and further directly transmitting the sensing data source to the data output layer for output. And further analyzing the first recognition target based on the output primary sensing data source and the secondary sensing data source, and sending an analysis result to an information terminal of a target user.
Further, as shown in fig. 3, step S500 of the present application further includes:
step S510: acquiring a plurality of sensors corresponding to the sensor network layer;
step S520: classifying based on the sensing data types of the plurality of sensors, and outputting a sensor classification result, wherein the data types output by each type of sensors are the same;
step S530: configuring preset characteristic intensity according to the sensor classification result;
step S540: and comparing the preset characteristic intensity with the plurality of data characteristic intensities to obtain the plurality of secondary sensing data sources.
Specifically, the sensor network layer is associated with a plurality of sensors for acquiring multi-type sensing data. The plurality of sensors are classified based on the collected sensing data types, and the sensors are classified based on the classification results of the sensing data types, wherein signals are output in a 0 and 1 mode or infrared sensing image output mode, and the classification results of the sensors are obtained. And respectively configuring the preset characteristic intensities according to the sensor classification result, namely judging whether the data characteristic intensity meets the critical characteristic intensity, wherein the preset characteristic intensities are in one-to-one correspondence with the sensor classification result due to different characteristic intensities of different sensing data types meeting the identification requirement. Mapping and corresponding are carried out on the plurality of preset characteristic intensities and the plurality of data characteristic intensities, whether the corresponding preset characteristic intensities are met or not is judged, when the preset characteristic intensities are not met, the fact that the corresponding primary sensing data sources are low in intrinsic characteristic intensity is indicated, the accuracy of data identification can be affected to a certain extent, sensing data acquisition needs to be carried out again, the secondary acquired sensing data sources are used as the plurality of secondary sensing data sources, so that the optimal acquisition of the characteristics of each group of data is guaranteed, and the accuracy of data identification is further improved.
Further, the step S540 of the present application further includes:
step S541: comparing the preset characteristic intensity with the plurality of data characteristic intensities to obtain N sensor data sources with the characteristic intensities smaller than the preset characteristic intensity;
step S542: the N sensor data sources are sent to N sensors correspondingly connected with the sensor network layer, and N data acquisition instructions are generated;
step S543: and controlling the N sensors to acquire data by the N data acquisition instructions to acquire the plurality of secondary sensing data sources.
Further, the generating N data acquisition instructions, step S542 of the present application further includes:
step S5421: acquiring N characteristic intensity differences corresponding to the N sensor data sources;
step S5422: and converting the N characteristic intensity differences to generate preset data acquisition quantity, and adding the preset data acquisition quantity into the N data acquisition instructions.
Specifically, the preset characteristic intensity is compared with the plurality of data characteristic intensities, the characteristic intensities smaller than the preset characteristic intensity in the plurality of data characteristic intensities are screened and extracted, the N sensor data sources, namely the data sources with weaker characteristic intensities and data secondary acquisition necessity are selected, the characteristic intensity differences are calculated respectively, and the N sensor data sources are marked. And sending the N sensor data sources to the sensor network layer, feeding back the N sensor data sources to the corresponding N sensors, and respectively generating N data acquisition instructions, namely, a start instruction for secondary acquisition of the data sources.
Specifically, based on the N sensor data sources, the N marked characteristic intensity differences are respectively extracted. The conversion ratio of the unit characteristic intensity difference to the data acquisition amount is configured, the conversion ratio is used as a measurement conversion standard, the data acquisition amount conversion is carried out on the N characteristic intensity differences, and the preset data acquisition amount is obtained, wherein the preset data acquisition amount is provided with a sensor identifier. And adding the preset data acquisition amount to the N corresponding data acquisition instructions. And along with the receiving of the N data acquisition instructions by the N sensors, taking the preset data acquisition instructions and the preset data acquisition instructions as acquisition standards, controlling the N sensors to acquire data, and taking the secondarily acquired sensing data as a plurality of secondary sensing data sources. By means of data characteristic intensity calibration, the data source with the necessity of secondary acquisition is accurately judged, the characteristic intensity difference is used as a measurement standard of the secondary acquisition quantity value of the data, and the acquisition quantity is accurately measured on the basis of guaranteeing the data identification requirement.
Further, the present application also includes step S700, including:
step S710: performing data fitting on the primary sensing data source and the secondary sensing data source to obtain fitting sensing data;
step S720: analyzing the first recognition target by using the fitting sensing data to obtain a recognition result of the first recognition target;
step S730: and transmitting the identification result to an information terminal of the target user by the sensor data sharing platform.
Specifically, in the primary sensing data source, data position positioning is performed on the secondary sensing data source, further data fitting is performed to complete summarization and normalization of the sensing data source, fitting sensing data is generated, the fitting sensing data is acquired data meeting data identification requirements, and each group of data is ensured to reach optimal acquisition. The fitting sensing data are sensing data of the first recognition target, multidimensional analysis such as state, rule and the like is carried out on the first recognition target, and an analysis recognition result is used as a recognition result of the first recognition target. And further taking the sensor data sharing platform as an intermediate medium, transmitting the identification result to an information terminal of a prime number target user, and directly acquiring the association information of the first identification target by the target user based on the identification result received by the information terminal.
Example two
Based on the same inventive concept as the intelligent recognition method based on multi-sensor data in the foregoing embodiments, as shown in fig. 4, the present application provides an intelligent recognition system based on multi-sensor data, the system comprising:
the platform building module 11 is used for building the sensor data sharing platform, wherein the sensor data sharing platform comprises a sensor network layer, a data transmission layer, a feature identification layer and a data output layer;
an information acquisition module 12, wherein the information acquisition module 12 is configured to acquire information of a first recognition target;
a primary sensing data source obtaining module 13, where the primary sensing data source obtaining module 13 is configured to input information of the first recognition target into the sensor data sharing platform, and obtain, according to the sensor network layer of the sensor data sharing platform, a plurality of primary sensing data sources based on the first recognition target, where the plurality of primary sensing data sources are in one-to-one correspondence with a plurality of sensors;
the characteristic intensity recognition module 14 is configured to transmit the plurality of primary sensing data sources to the characteristic recognition layer according to the data transmission layer to perform data characteristic intensity recognition, so as to obtain a plurality of data characteristic intensities;
the secondary sensing data source acquisition module 15 is configured to acquire a plurality of secondary sensing data sources according to the plurality of data feature intensities;
the data source output module 16 is configured to output the primary sensing data source and the secondary sensing data source through the data output layer by using the data source output module 16.
Further, the system further comprises:
the sensor acquisition module is used for acquiring a plurality of sensors corresponding to the sensor network layer;
the sensor classification module classifies the sensor data types based on the sensors and outputs a sensor classification result, wherein the data types output by each type of sensor are the same;
the intensity configuration module is used for configuring preset characteristic intensity according to the sensor classification result;
and the intensity comparison module is used for comparing the preset characteristic intensity with the plurality of data characteristic intensities to obtain the plurality of secondary sensing data sources.
Further, the system further comprises:
the data source acquisition module is used for comparing the preset characteristic intensity with the plurality of data characteristic intensities to acquire N sensor data sources with the characteristic intensities smaller than the preset characteristic intensity;
the instruction generation module is used for sending the N sensor data sources to N sensors correspondingly connected with the sensor network layer to generate N data acquisition instructions;
the data acquisition module is used for controlling the N sensors to acquire data according to the N data acquisition instructions, and acquiring the plurality of secondary sensing data sources.
Further, the system further comprises:
the characteristic intensity difference acquisition module is used for acquiring N characteristic intensity differences corresponding to the N sensor data sources;
the acquisition amount adding module is used for converting the N characteristic intensity differences to generate preset data acquisition amounts and adding the preset data acquisition amounts into the N data acquisition instructions.
Further, the system further comprises:
the data fitting module is used for carrying out data fitting on the primary sensing data source and the secondary sensing data source to obtain fitting sensing data;
the target analysis module is used for analyzing the first identification target by the fitting sensing data to acquire an identification result of the first identification target;
and the result transmission module is used for transmitting the identification result to the information terminal of the target user through the sensor data sharing platform.
Further, the system further comprises:
the data source transmission module is used for transmitting the primary sensing data source to the characteristic recognition layer, wherein the characteristic recognition layer is embedded with a characteristic intensity recognition model;
the characteristic parameter acquisition module is used for analyzing according to the characteristic intensity recognition model to acquire characteristic data sensing granularity, characteristic data time sequence and characteristic data labeling quantity;
and the characteristic intensity calculation module is used for calculating characteristic intensity weights by taking the characteristic data sensing granularity, the characteristic data time sequence and the characteristic data labeling quantity as variables and outputting the plurality of data characteristic intensities.
Further, the system further comprises:
the sample acquisition module is used for acquiring a test sample data set;
the sample transmission analysis module is used for inputting the test sample data set into the data transmission layer for analysis to obtain a transmission sample data set;
the transmission loss acquisition module is used for comparing the transmission sample data set with the test sample data set to obtain a transmission loss coefficient;
and the data characteristic intensity adjusting module is used for adjusting the plurality of data characteristic intensities according to the transmission loss coefficient.
In the present disclosure, through the foregoing detailed description of a method for intelligent recognition based on multi-sensor data, those skilled in the art may clearly know a method and a system for intelligent recognition based on multi-sensor data in this embodiment, and for a device disclosed in the embodiment, since the device corresponds to a method disclosed in the embodiment, the description is relatively simple, and relevant places refer to the description of the method section.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (3)
1. An intelligent recognition method based on multi-sensor data, wherein the method is applied to a multi-sensor data processing system, the system is in communication connection with a sensor data sharing platform, and the method comprises the following steps:
the sensor data sharing platform is built, wherein the sensor data sharing platform comprises a sensor network layer, a data transmission layer, a feature identification layer and a data output layer;
acquiring information of a first identification target;
inputting the information of the first recognition target into the sensor data sharing platform, and acquiring a plurality of primary sensing data sources based on the first recognition target according to the sensor network layer of the sensor data sharing platform, wherein the plurality of primary sensing data sources are in one-to-one correspondence with a plurality of sensors;
transmitting the plurality of primary sensing data sources to the characteristic recognition layer according to the data transmission layer to recognize the data characteristic intensity, and acquiring a plurality of data characteristic intensities;
acquiring a plurality of secondary sensing data sources according to the plurality of data characteristic intensities;
outputting the primary sensing data source and the secondary sensing data source through the data output layer;
transmitting the plurality of primary sensing data sources to the feature recognition layer for data feature intensity recognition, wherein the method comprises the following steps of:
transmitting the primary sensing data source to the feature recognition layer, wherein the feature recognition layer is embedded with a feature intensity recognition model;
analyzing according to the characteristic intensity recognition model to obtain characteristic data sensing granularity, characteristic data chronology and characteristic data labeling quantity;
performing feature intensity weight calculation by taking the feature data sensing granularity, the feature data time sequence and the feature data labeling quantity as variables, and outputting the plurality of data feature intensities;
the feature data sensing granularity is an extraction node of a feature data sensing data source, the feature data time sequence is a sequence of feature extraction time nodes, and the feature data labeling amount is a manual labeling existing after the primary sensing data source is initially acquired;
acquiring a test sample data set;
inputting the test sample data set into the data transmission layer for analysis to obtain a transmission sample data set;
comparing the transmission sample data set with the test sample data set to obtain a transmission loss coefficient;
adjusting the characteristic intensities of the data according to the transmission loss coefficient;
the method further comprises the steps of:
acquiring a plurality of sensors corresponding to the sensor network layer;
classifying based on the sensing data types of the plurality of sensors, and outputting a sensor classification result, wherein the data types output by each type of sensors are the same;
configuring preset characteristic intensity according to the sensor classification result;
comparing the preset characteristic intensity with the plurality of data characteristic intensities to obtain a plurality of secondary sensing data sources;
the method for acquiring a plurality of secondary sensing data sources comprises the following steps:
comparing the preset characteristic intensity with the plurality of data characteristic intensities to obtain N sensor data sources with the characteristic intensities smaller than the preset characteristic intensity;
the N sensor data sources are sent to N sensors correspondingly connected with the sensor network layer, and N data acquisition instructions are generated;
the N sensors are controlled to acquire data by the N data acquisition instructions, and the plurality of secondary sensing data sources are acquired;
the generating N data acquisition instructions further includes:
acquiring N characteristic intensity differences corresponding to the N sensor data sources;
and converting the N characteristic intensity differences to generate preset data acquisition quantity, and adding the preset data acquisition quantity into the N data acquisition instructions.
2. The method of claim 1, wherein the method further comprises:
performing data fitting on the primary sensing data source and the secondary sensing data source to obtain fitting sensing data;
analyzing the first recognition target by using the fitting sensing data to obtain a recognition result of the first recognition target;
and transmitting the identification result to an information terminal of the target user by the sensor data sharing platform.
3. An intelligent recognition system based on multi-sensor data, wherein the system is communicatively connected to a sensor data sharing platform, the system comprising:
the platform building module is used for building the sensor data sharing platform, wherein the sensor data sharing platform comprises a sensor network layer, a data transmission layer, a feature identification layer and a data output layer;
the information acquisition module is used for acquiring information of the first identification target;
the primary sensing data source acquisition module is used for inputting the information of the first identification target into the sensor data sharing platform, and acquiring a plurality of primary sensing data sources based on the first identification target according to the sensor network layer of the sensor data sharing platform, wherein the plurality of primary sensing data sources are in one-to-one correspondence with a plurality of sensors;
the characteristic intensity recognition module is used for transmitting the plurality of primary sensing data sources to the characteristic recognition layer to recognize the characteristic intensity of the data according to the data transmission layer, and acquiring a plurality of characteristic intensities of the data;
the secondary sensing data source acquisition module is used for acquiring a plurality of secondary sensing data sources according to the plurality of data characteristic intensities;
the data source output module is used for outputting the primary sensing data source and the secondary sensing data source through the data output layer;
the data source transmission module is used for transmitting the primary sensing data source to the characteristic recognition layer, wherein the characteristic recognition layer is embedded with a characteristic intensity recognition model;
the characteristic parameter acquisition module is used for analyzing according to the characteristic intensity recognition model to acquire characteristic data sensing granularity, characteristic data time sequence and characteristic data labeling quantity;
the characteristic intensity calculation module is used for calculating characteristic intensity weights by taking the characteristic data sensing granularity, the characteristic data time sequence and the characteristic data labeling quantity as variables and outputting the plurality of data characteristic intensities, wherein the characteristic data sensing granularity is an extraction node of a characteristic data sensing data source, the characteristic data time sequence is a sequence of characteristic extraction time nodes, and the characteristic data labeling quantity is a manual label existing after primary acquisition of the sensing data source;
the sample acquisition module is used for acquiring a test sample data set;
the sample transmission analysis module is used for inputting the test sample data set into the data transmission layer for analysis to obtain a transmission sample data set;
the transmission loss acquisition module is used for comparing the transmission sample data set with the test sample data set to obtain a transmission loss coefficient;
the data characteristic intensity adjusting module is used for adjusting the plurality of data characteristic intensities according to the transmission loss coefficient;
the sensor acquisition module is used for acquiring a plurality of sensors corresponding to the sensor network layer;
the sensor classification module classifies the sensor data types based on the sensors and outputs a sensor classification result, wherein the data types output by each type of sensor are the same;
the intensity configuration module is used for configuring preset characteristic intensity according to the sensor classification result;
the intensity comparison module is used for comparing the preset characteristic intensity with the plurality of data characteristic intensities to obtain a plurality of secondary sensing data sources;
the data source acquisition module is used for comparing the preset characteristic intensity with the plurality of data characteristic intensities to acquire N sensor data sources with the characteristic intensities smaller than the preset characteristic intensity;
the instruction generation module is used for sending the N sensor data sources to N sensors correspondingly connected with the sensor network layer to generate N data acquisition instructions;
the data acquisition module is used for controlling the N sensors to acquire data according to the N data acquisition instructions to acquire the plurality of secondary sensing data sources;
the characteristic intensity difference acquisition module is used for acquiring N characteristic intensity differences corresponding to the N sensor data sources; the acquisition amount adding module is used for converting the N characteristic intensity differences to generate preset data acquisition amounts and adding the preset data acquisition amounts into the N data acquisition instructions.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310254438.1A CN115964678B (en) | 2023-03-16 | 2023-03-16 | Intelligent identification method and system based on multi-sensor data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310254438.1A CN115964678B (en) | 2023-03-16 | 2023-03-16 | Intelligent identification method and system based on multi-sensor data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115964678A CN115964678A (en) | 2023-04-14 |
| CN115964678B true CN115964678B (en) | 2023-10-03 |
Family
ID=87354984
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310254438.1A Active CN115964678B (en) | 2023-03-16 | 2023-03-16 | Intelligent identification method and system based on multi-sensor data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115964678B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106897839A (en) * | 2017-03-08 | 2017-06-27 | 武汉盛华伟业科技有限公司 | A kind of data integrated collaboration platform in well site |
| CN110664412A (en) * | 2019-09-19 | 2020-01-10 | 天津师范大学 | Human activity recognition method facing wearable sensor |
| EP3816879A1 (en) * | 2019-11-04 | 2021-05-05 | Gaf AG | A method of yield estimation for arable crops and grasslands and a system for performing the method |
| WO2021139424A1 (en) * | 2020-05-14 | 2021-07-15 | 平安科技（深圳）有限公司 | Text content quality evaluation method, apparatus and device, and storage medium |
| CN114330318A (en) * | 2021-12-20 | 2022-04-12 | 华南师范大学 | Method and device for recognizing Chinese fine-grained entities in financial field |
| CN114973120A (en) * | 2022-04-14 | 2022-08-30 | 山东大学 | Behavior identification method and system based on multi-dimensional sensing data and monitoring video multi-mode heterogeneous fusion |
| CN115292285A (en) * | 2022-08-04 | 2022-11-04 | 中国人民解放军国防科技大学 | Distributed architecture-based data topic management method and system |
| CN115661376A (en) * | 2022-12-28 | 2023-01-31 | 深圳市安泽拉科技有限公司 | Target reconstruction method and system based on unmanned aerial vehicle image |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| ES2964705T3 (en) * | 2016-05-06 | 2024-04-09 | Univ Leland Stanford Junior | Mobile and portable video capture and feedback platforms for the therapy of mental disorders |
| US20210382893A1 (en) * | 2020-06-08 | 2021-12-09 | Dropbox, Inc. | Intelligently identifying the most knowledgable person based on multiple data source inputs |
-
2023
- 2023-03-16 CN CN202310254438.1A patent/CN115964678B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106897839A (en) * | 2017-03-08 | 2017-06-27 | 武汉盛华伟业科技有限公司 | A kind of data integrated collaboration platform in well site |
| CN110664412A (en) * | 2019-09-19 | 2020-01-10 | 天津师范大学 | Human activity recognition method facing wearable sensor |
| EP3816879A1 (en) * | 2019-11-04 | 2021-05-05 | Gaf AG | A method of yield estimation for arable crops and grasslands and a system for performing the method |
| WO2021139424A1 (en) * | 2020-05-14 | 2021-07-15 | 平安科技（深圳）有限公司 | Text content quality evaluation method, apparatus and device, and storage medium |
| CN114330318A (en) * | 2021-12-20 | 2022-04-12 | 华南师范大学 | Method and device for recognizing Chinese fine-grained entities in financial field |
| CN114973120A (en) * | 2022-04-14 | 2022-08-30 | 山东大学 | Behavior identification method and system based on multi-dimensional sensing data and monitoring video multi-mode heterogeneous fusion |
| CN115292285A (en) * | 2022-08-04 | 2022-11-04 | 中国人民解放军国防科技大学 | Distributed architecture-based data topic management method and system |
| CN115661376A (en) * | 2022-12-28 | 2023-01-31 | 深圳市安泽拉科技有限公司 | Target reconstruction method and system based on unmanned aerial vehicle image |
Non-Patent Citations (2)
| Title |
|---|
| 王永利 ; 刘景春 ; 于占龙 ; 富春岩 ; 刁树民 ; .传感数据起源信息的多粒度标注与多分辨压缩.南通大学学报(自然科学版).2010,(第03期),全文. * |
| 郝志峰 ; 黄浩 ; 蔡瑞初 ; 温雯 ; .基于多特征融合与双向RNN的细粒度意见分析.计算机工程.2018,(第07期),全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115964678A (en) | 2023-04-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109064454A (en) | Product defects detection method and system | |
| CN110411580B (en) | Diagnosis method and diagnosis system for heating defects of electric equipment | |
| CN111353379B (en) | Signal measurement feature matching and labeling method based on weight clustering | |
| CN107423754B (en) | Automatic radiation source identification system based on parameter multi-attribute autonomous intelligent decision | |
| CN105022792B (en) | Passive radar signal sorting associated weights calculation method based on data mining | |
| CN108537790B (en) | Different-source image change detection method based on coupling translation network | |
| CN112181006B (en) | Environment intelligent processing method and system based on big data and storage medium | |
| CN116092214B (en) | Synchronous monitoring method and system for production of lightweight body-in-white assembly | |
| CN114994261A (en) | Remote calibration method for sensor for ambient air monitoring | |
| CN109375948A (en) | A kind of software pricing method of intelligent function point identification | |
| CN116600104B (en) | Phase acquisition quality analysis method and system for IPC network camera | |
| CN111582387A (en) | Rock spectral feature fusion classification method and system | |
| CN116228028B (en) | Application performance evaluation method and system for plastic bags | |
| CN115964678B (en) | Intelligent identification method and system based on multi-sensor data | |
| CN110072205A (en) | A kind of layering aggregation method for wireless sense network anomaly data detection | |
| CN114861705A (en) | Electromagnetic target intelligent perception identification method based on multi-feature heterogeneous fusion | |
| CN111967501B (en) | Method and system for judging load state driven by telemetering original data | |
| CN117037771B (en) | Zero cold water control method and device based on voice recognition storage | |
| CN116859875B (en) | Steel pipe production process adjusting and controlling method and system based on use requirements | |
| CN116152709A (en) | Intelligent grading pretreatment method and system for decoration garbage | |
| CN115150430A (en) | Automatic vending machine operation data acquisition system based on thing networking | |
| CN116071532B (en) | Intelligent door control method and system for outdoor cabinet | |
| CN115039549A (en) | Crop planting and fertilizing information processing method and system | |
| CN113884756A (en) | Electric energy metering edge acquisition device and method | |
| CN113128158A (en) | Coupling coefficient identification method of wireless power transmission system |
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 |