阅读说明：本技术 一种大数据智能探测设备辅助增强系统 (Auxiliary enhancement system for big data intelligent detection equipment ) 是由 不公告发明人 于 2020-06-29 设计创作，主要内容包括：本发明公开了一种大数据智能探测设备辅助增强系统,涉及人工智能领域,系统包括：离线数据研究中心和在线实时处理系统。离线数据研究中心用于收集存储和管理各型号设备的历史使用数据,并提供数据检索服务；离线数据研究中心还用于执行算法训练更新与研发任务,训练更新已有业务算法模型和研发新的业务算法模型。在线实时处理系统通过接入预设设备产生的探测数据,利用预设设备对应的业务算法模型计算产生计算结果,在线实时处理系统将其计算结果与输入的数据导出送回至离线数据研究中心存储及用于业务算法模型的更新或迭代。本发明实现了算法模型能够较为容易的在设备上进行更新和替换,实现了快速、高效地将新的模型应用到设备中。(The invention discloses an auxiliary enhancement system of big data intelligent detection equipment, which relates to the field of artificial intelligence and comprises the following components: an off-line data research center and an on-line real-time processing system. The off-line data research center is used for collecting, storing and managing historical use data of various types of equipment and providing data retrieval service; the off-line data research center is also used for executing algorithm training updating and research and development tasks, training and updating the existing business algorithm model and researching and developing a new business algorithm model. The online real-time processing system generates a calculation result by accessing detection data generated by preset equipment and utilizing a business algorithm model corresponding to the preset equipment for calculation, and the online real-time processing system exports the calculation result and input data and sends the calculation result and the input data back to the offline data research center for storage and is used for updating or iterating the business algorithm model. The invention realizes that the algorithm model can be updated and replaced on the equipment more easily, and realizes that the new model is applied to the equipment quickly and efficiently.)

1. A big data intelligent detection device auxiliary enhancement system, characterized in that, the system includes:

the system comprises an offline data research center and an online real-time processing system;

the off-line data research center is used for collecting historical use data of various types of equipment, storing and managing the collected historical use data and providing data retrieval service for research and development of a business algorithm model; the off-line data research center is also used for executing algorithm training updating and research and development tasks, training and updating the existing business algorithm model and researching and developing a new business algorithm model according to real-time data in the off-line data research center, and the off-line data research center is used for providing support for the application of the business algorithm model at the equipment end;

the online real-time processing system is built at each equipment end, a business algorithm model in the offline data research center is applied to the equipment end through the online real-time processing system, the online real-time processing system generates a calculation result through accessing detection data generated by preset equipment and calculating by using the business algorithm model corresponding to the preset equipment, the calculation result is displayed through a visual display and control interface, and the online real-time processing system leads out the calculation result and input data and sends the calculation result and the input data back to the offline data research center for storage and is used for updating or iterating the business algorithm model.

2. The big data intelligent detection equipment auxiliary enhancement system according to claim 1, characterized in that hardware in the offline data research center comprises a plurality of server clusters, signal transmission equipment, an operation and maintenance control terminal and a data service application terminal, the server clusters are connected with the operation and maintenance control terminal and the data service application terminal through the signal transmission equipment, the plurality of server clusters are used for realizing big data distributed storage and parallelization calculation, and the operation and maintenance control terminal and the data service application terminal are used for realizing operation maintenance and data analysis mining of the offline data research center.

3. The big data intelligent detection equipment auxiliary enhancement system according to claim 1, wherein hardware in the online real-time processing system comprises a server cabinet and a display unit, the server cabinet is connected with the display unit, and the server cabinet comprises: the server cabinet is connected with the intelligent detection equipment through the signal transmission equipment, receives data generated by detection of the intelligent detection equipment, calculates based on a service algorithm model in the server cabinet and the data generated by detection to obtain an analysis result, and displays the analysis result in the display unit.

4. The big data intelligent probe device auxiliary enhancement system of claim 1, wherein the software in the offline data research center comprises: the system comprises a big data base platform, a big data management system, a big data analysis and mining system and a big data visualization system;

the big data base platform provides distributed data storage, a parallel computing engine, a distributed resource management system, a distributed application coordination service and a platform operation and maintenance management system, and is used for supporting other big data application systems based on the platform;

the big data management system is used for importing and transferring mass detection data, carrying out unified storage management on the mass detection data and efficiently and flexibly retrieving, inquiring and using the mass detection data;

the big data analysis and mining system is used for analyzing, mining and modeling massive structured data, analyzing and mining hidden information and knowledge from the massive data, improving the efficiency of data value discovery and realizing data-driven business innovation;

and the big data visualization system is used for carrying out service evaluation and customized comprehensive detection situation display on the modeling result generated by the big data analysis and mining system.

5. The big data intelligent detecting equipment auxiliary enhancement system according to claim 1, wherein the software in the online real-time processing system comprises:

the data access acquisition module is used for receiving real-time detection data generated by the intelligent detection equipment;

the model calling calculation analysis module is used for calling a built-in service algorithm model to realize corresponding detection service;

the model and data management module is used for managing all business algorithm models, detection data and calculation results in the online real-time processing system;

the display control interaction module is used for connecting the display unit and realizing the online real-time processing system service display and the corresponding human-computer interaction function;

and the storage and resource management module is used for managing, distributing and scheduling the storage resources and the computing resources in the online real-time processing system.

6. The big data intelligent detection equipment auxiliary enhancement system according to claim 1, wherein the development and application process of the business algorithm model is as follows:

collecting and labeling detection data of the intelligent detection equipment, and uniformly storing and managing the detection data by an offline data research center;

the off-line data research center builds, trains and verifies a business algorithm model based on data in the center;

exporting the verified business algorithm model from an offline data research center, and importing the business algorithm model into an online real-time processing system;

the on-line real-time processing system applies the service algorithm model to complete the corresponding service.

7. The big data intelligent detection device auxiliary enhancement system according to claim 1, wherein the auxiliary enhancement system is used by the following procedures:

building an auxiliary enhancement system, researching and developing a business algorithm model and verifying an offline business algorithm model in an offline data research center after the building is finished, detecting historical data by using intelligent detection equipment stored in the offline data research center, building the business algorithm model, verifying the business algorithm model in an offline manner, and exporting the business algorithm model after verification;

after the business algorithm model is exported, the business algorithm model is sent to an online real-time processing system at the equipment end, the business algorithm model is uploaded to the online real-time processing system, the online real-time processing system is accessed into intelligent detection equipment data, the business algorithm model is called to analyze and process the real-time detection data, and the calculation result of the business algorithm model is displayed through a display unit to complete the corresponding business function;

storing the calculation result of the business algorithm model and the input data, exporting the stored data and sending the exported data back to the offline data research center, and updating or iterating the business algorithm model;

the off-line data research center updates the stored related data of the intelligent detection equipment, and develops a new business algorithm model or updates or iterates the business algorithm model being used based on the updated data.

8. The big data smart probe device aiding enhancement system of claim 1, wherein the smart probe device is a radar or sonar probe device.

9. The big data intelligent detection device auxiliary enhancement system according to claim 2, wherein the server is an x86 server, and the signal transmission device is a switch.

10. The big data intelligent detection device auxiliary enhancement system of claim 8, wherein the business algorithm model is an interference suppression model or an object type recognition model.

Technical Field

The invention relates to the field of artificial intelligence, in particular to an auxiliary enhancement system of big data intelligent detection equipment.

Background

Radars use modulated waveforms and directional antennas to transmit electromagnetic energy into a particular area of space to search for targets. Objects (targets) within the search area reflect some of the energy (radar reflection or echoes) back to the radar, and these echoes are then processed by the radar receiver to extract information about the target such as range, velocity, angular position, and other target identification features. The sonar also works in a similar way, except that the sonar utilizes the propagation characteristic of sound signals under water to transmit or receive underwater sound wave signals, and then completes the conversion between sound waves (mechanical waves) and electromagnetic waves through a transducer so as to extract target identification features. According to the detection principle of radar and sonar, extracting the relevant information of the target from the signal waveform received by the equipment is a necessary and key step in the detection process. At present, the most common method is to adopt a technical means in the aspect of Digital Signal Processing (DSP) to perform signal noise reduction, target extraction, target spectrum analysis, feature extraction and other processing processes on a received detection signal, and to visually and audibly present a processing result in real time, so that a user can find, identify and track a target in time.

Aiming at the challenges and difficulties faced by radar and sonar detection, a new solution idea and method is as follows: for signal data of a detection target, on the basis of a digital signal processing method, a deep learning technology in big data mining is utilized, and a target classification deep neural network is trained in a data driving mode and used for realizing classification and identification of the detection target.

Based on mass actual detection data, a deep learning technology is used, and time-frequency signal data of a target are directly used as training input. Compared with the traditional type identification based on theoretical characteristics, the method eliminates the limitation of characteristic extraction on classification models, solves the problem from the characteristics of data, trains different models to correspond to different data conditions according to different environments, and has higher flexibility and universality compared with the traditional method.

However, applying a deep learning model based on data driving on an actual detection device also faces a great challenge. Firstly, a traditional equipment development mode follows a route of algorithm simulation- > hardware system design and construction- > auxiliary control software construction, related business algorithm software is compiled into embedded hardware or chips such as an FPGA through a bottom layer, and rewriting and compiling of the bottom layer algorithm on the hardware consumes more time and is more difficult compared with conventional software coding, especially on equipment which is produced and shaped, a series of problems of equipment disassembly and assembly, reduction and the like are involved, so that in the existing application mode, the algorithm model on the equipment is not easy to update and replace, and the efficiency of applying a new model to the equipment is low.

Disclosure of Invention

The invention provides an auxiliary enhancement system for big data intelligent detection equipment, and aims to realize that an algorithm model can be updated and replaced on the equipment easily and to apply a new model to the equipment quickly and efficiently.

In order to achieve the above object, the present invention provides an auxiliary enhancement system for big data intelligent detection equipment, wherein the system comprises:

the system comprises an offline data research center and an online real-time processing system;

the off-line data research center is used for collecting historical use data of various types of equipment, storing and managing the collected historical use data and providing data retrieval service for research and development of a business algorithm model; the off-line data research center is also used for executing algorithm training updating and research and development tasks, training and updating the existing business algorithm model and researching and developing a new business algorithm model according to real-time data in the off-line data research center, and the off-line data research center is used for providing support for the application of the business algorithm model at the equipment end;

the online real-time processing system is built at each equipment end, a business algorithm model in the offline data research center is applied to the equipment end through the online real-time processing system, the online real-time processing system generates a calculation result through accessing detection data generated by preset equipment and calculating by using the business algorithm model corresponding to the preset equipment, the calculation result is displayed through a visual display and control interface, and the online real-time processing system leads out the calculation result and input data and sends the calculation result and the input data back to the offline data research center for storage and is used for updating or iterating the business algorithm model.

The principle of the invention is as follows: the deep learning model has the advantages that retraining iteration can be carried out on the model along with the change of data, so that the performance of the model can be kept at a certain level; secondly, the deep learning model is developed faster than the traditional processing algorithm, and new model designs with better performance are continuously published, so that how to quickly and efficiently apply the new model to the equipment is a pain point which is generally and urgently needed to be solved by the traditional equipment. The invention is based on the existing in-use equipment, builds a corresponding intelligent detection auxiliary enhancement system in a matching way, connects the system with the equipment in the form of external auxiliary equipment, receives real-time data, and then generates a recognition result through calculation of a core business algorithm model. In addition, a corresponding offline data storage and processing center is required to be built for accumulating and storing historical data, continuously iterating and optimizing the lifting model, and updating the lifting model into an external system. The invention separates the operation of the service algorithm from the hardware system of the equipment, and realizes the calculation of the service algorithm in the system by accessing the data stream generated by the detection equipment into the system of the invention. The subsequent algorithm updating only needs to be carried out by using a special software module in the system, the original equipment does not need to be unpacked, and the reprogramming development on the bottom embedded chip is also not needed, so that the algorithm model can be updated and replaced on the equipment easily, and the new model can be applied to the equipment quickly and efficiently.

The invention can provide a feasible route for developing new equipment on one hand, and can provide a set of practical and effective method for upgrading and reconstructing old equipment and improving performance on the other hand.

The intelligent detection auxiliary enhancement system utilizes big data artificial intelligence theory and technical means to build an algorithm model and an application system, is used for assisting detection equipment to quickly and accurately find a target, and improves the comprehensive detection capability of the existing detection equipment such as radar, sonar and the like. The system constructs detection-related business algorithm models (an interference suppression model, a target type identification model and the like) by collecting detection historical data of corresponding equipment on the basis of not changing the structure of original equipment, and is externally connected to detection equipment in an application system mode, so that real-time system application is realized, and intelligent auxiliary detection is carried out.

Preferably, hardware in the offline data research center comprises a plurality of server clusters, signal transmission equipment, an operation and maintenance control terminal and a data service application terminal, the server clusters are connected with the operation and maintenance control terminal and the data service application terminal through the signal transmission equipment, the server clusters are used for achieving distributed storage and parallelization calculation of big data, and the operation and maintenance control terminal and the data service application terminal are used for achieving operation maintenance and data analysis and mining of the offline data research center.

Preferably, the hardware in the online real-time processing system includes a server cabinet and a display unit, the server cabinet is connected to the display unit, and the server cabinet includes: the server cabinet is connected with the intelligent detection equipment through the signal transmission equipment, receives data generated by detection of the intelligent detection equipment, calculates based on a service algorithm model in the server cabinet and the data generated by detection to obtain an analysis result, and displays the analysis result in the display unit.

Preferably, the software in the off-line data research center includes: the system comprises a big data base platform, a big data management system, a big data analysis and mining system and a big data visualization system;

the big data base platform provides distributed data storage, a parallel computing engine, a distributed resource management system, a distributed application coordination service and a platform operation and maintenance management system, and is used for supporting other big data application systems based on the platform;

the big data management system is used for importing and transferring mass detection data, carrying out unified storage management on the mass detection data and efficiently and flexibly retrieving, inquiring and using the mass detection data;

the big data analysis and mining system is used for analyzing, mining and modeling massive structured data, analyzing and mining hidden information and knowledge from the massive data, improving the efficiency of data value discovery and realizing data-driven business innovation;

and the big data visualization system is used for carrying out service evaluation and customized comprehensive detection situation display on the modeling result generated by the big data analysis and mining system.

Preferably, the software in the online real-time processing system comprises:

the data access acquisition module is used for receiving real-time detection data generated by the intelligent detection equipment;

the model calling calculation analysis module is used for calling a built-in service algorithm model to realize corresponding detection service;

the model and data management module is used for managing all business algorithm models, detection data and calculation results in the online real-time processing system;

the display control interaction module is used for connecting the display unit and realizing the online real-time processing system service display and the corresponding human-computer interaction function;

and the storage and resource management module is used for managing, distributing and scheduling the storage resources and the computing resources in the online real-time processing system.

Preferably, the research and development and application process of the business algorithm model is as follows:

collecting and labeling detection data of the intelligent detection equipment, and uniformly storing and managing the detection data by an offline data research center;

the off-line data research center builds, trains and verifies a business algorithm model based on data in the center;

exporting the verified business algorithm model from an offline data research center, and importing the business algorithm model into an online real-time processing system;

and the online real-time processing system is applied to the business algorithm model to complete corresponding business.

Preferably, the usage process of the auxiliary enhancement system is as follows:

building an auxiliary enhancement system, researching and developing and verifying a business algorithm model in an offline data research center after the building is finished, detecting historical data by using intelligent detection equipment stored in the offline data research center, building the business algorithm model, verifying the business algorithm model in an offline manner, and exporting the business algorithm model after verification;

after the business algorithm model is exported, the business algorithm model is sent to an online real-time processing system at the equipment end, the business algorithm model is uploaded to the online real-time processing system, the online real-time processing system is accessed into intelligent detection equipment data, the business algorithm model is called to analyze and process the real-time detection data, and the calculation result of the business algorithm model is displayed through a display unit to complete the corresponding business function;

storing the calculation result of the business algorithm model and the input data, exporting the stored data and sending the exported data back to the offline data research center, and updating or iterating the business algorithm model;

the off-line data research center updates the stored related data of the intelligent detection equipment, and develops a new business algorithm model or updates or iterates the business algorithm model being used based on the updated data.

Preferably, the intelligent detection device is a radar or sonar detection device.

Preferably, the server is an x86 server and the signaling device is a switch.

Preferably, the business algorithm model is an interference suppression model or an object type identification model.

One or more technical schemes provided by the invention at least have the following technical effects or advantages:

by adding the big data intelligent detection auxiliary enhancement system to the existing equipment, a large amount of time and resources required by developing new-type equipment are avoided, resources can be concentrated on research and development construction of the enhancement system, and an actual effective service model and a mature application mode are explored. The radar and sonar intelligent detection auxiliary enhancement system is built, on one hand, the intelligent degree and the actual detection capability of old equipment are improved, on the other hand, a large amount of practical experience is accumulated for the development of subsequent new equipment from system construction to core algorithm model construction, and the development of the equipment and the improvement of the detection efficiency are facilitated. The system can well apply the latest deep learning model on the old and new equipment, and can rapidly update the iterative model along with the change of environment, time and data, so that the detection effect is guaranteed.

Drawings

The accompanying drawings, which are included to provide a further understanding of the embodiments of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention;

FIG. 1 is a schematic diagram of a logic structure of a big data intelligent detection auxiliary enhancement system;

FIG. 2 is a schematic diagram of an off-line data research center hardware architecture;

FIG. 3 is a diagram of an on-line real-time processing system hardware architecture;

FIG. 4 is a schematic diagram of an off-line data research center software architecture;

FIG. 5 is a diagram of an on-line real-time processing system software architecture.

Detailed Description

In order that the above objects, features and advantages of the present invention can be more clearly understood, a more particular description of the invention will be rendered by reference to the appended drawings. It should be noted that the embodiments and features of the embodiments of the present application may be combined with each other without conflicting with each other.

In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, however, the present invention may be practiced in other ways than those specifically described and thus the scope of the present invention is not limited by the specific embodiments disclosed below.