CN115079609A

CN115079609A - System and method for managing personnel behaviors and equipment working condition behaviors based on multiple scenes of Internet of things

Info

Publication number: CN115079609A
Application number: CN202210734704.6A
Authority: CN
Inventors: 王智博; 王志健; 王斌
Original assignee: Shanghai Tuowang Data Technology Co ltd
Current assignee: Shanghai Tuowang Data Technology Co ltd
Priority date: 2022-06-27
Filing date: 2022-06-27
Publication date: 2022-09-20

Abstract

The invention provides a multi-scenario manager behavior and equipment working condition behavior system and a method based on the Internet of things, wherein the system comprises the following components: a data acquisition module, comprising: the system comprises an intelligent terminal, wireless charging equipment and a communication system; the intelligent terminal is used for monitoring, acquiring and transmitting personnel positioning control behavior data and equipment positioning control behavior data; the intelligent terminal includes: the device comprises a positioning module, a sensor module, an equipment sensing module, a power supply module and an abnormality detection processing module; the personnel management module is used for acquiring basic information of the user through management operation; the device management module is used for acquiring device use information through management operation; and the intelligent management platform is used for storing data and analyzing staff clients, acquiring physical condition information, giving out early warning when the detected staff is judged to be abnormal, carrying out intelligent management on the staff or equipment, and counting to generate a report. The invention solves the technical problems of dependence on manpower, higher management cost, lower management efficiency and lower operation precision.

Description

System and method for managing personnel behaviors and equipment working condition behaviors based on multiple scenes of Internet of things

Technical Field

The invention relates to the field of application of the Internet of things, in particular to a system and a method for managing personnel behaviors and equipment working condition behaviors based on multiple scenes of the Internet of things.

Background

With the continuous deepening of reform development and market economy, market competition is increasingly violent, in many industries, staff are high in mobility, capital circulation is high, management and dispute problems are frequent, and how to improve product quality and reduce internal consumption becomes an important link for survival and development of enterprises. Throughout the country and abroad, the scientific and technical content of products is greatly improved regardless of the size of enterprises, meanwhile, the cost is reduced by an internal positive method, good enterprise images and high-quality products can win customers, and a high-speed copying success mode is needed for the outside, so that the operation is expanded, the market is occupied, and a larger profit space is obtained. However, to ensure the smooth implementation of the method, a scientific processing method is adopted, advanced management concepts are also applied, the management method and regulations are related to subjective behaviors of managers, and are often different from person to person and different in space-time background, and particularly, the management of workers and equipment is enhanced, so that the method is closely related to comprehensive economic benefits and is also closely related to the healthy development of enterprises. At present, the use of the technology of the internet of things has spread in many scenes, but how to achieve comprehensive management and control of positioning, real-time monitoring and control of workers and equipment in an enterprise is a problem to be considered by related enterprises and departments.

The internet of things is a network which connects articles or people with the network through various sensing devices and carries out information communication so as to realize intelligent identification, positioning, tracking, monitoring and management. Through the Internet of things, an intelligent system which can be shared between people and objects at any time can be established in the fields of intelligent transportation, traditional industry, intelligent logistics, engineering control, intelligent medical treatment, city management, public safety, intelligent home, farming and animal husbandry production and the like. The Internet of things which intelligently connects people with objects and objects is a one-time historical opportunity that the Chinese information industry overtakes the world.

With the development of internet technology, internet of things technology is changing our lives. In the management of personnel and equipment of an enterprise, aiming at the characteristics of large mobility of personnel in operation of the enterprise, high capital circulation, various equipment types, complex equipment conditions, high equipment use frequency and the like, the conventional manual or manual management and maintenance mode is utilized, and good enterprise operation and control cost is difficult to guarantee.

Thereby, the following problems are generated: when an enterprise cannot monitor staff in real time, how to monitor the working state of the staff and evaluate the working efficiency is realized; when an enterprise cannot monitor equipment in real time, how to reduce illegal operation of equipment use is reduced, and the aim of reducing cost is fulfilled. Therefore, monitoring and managing the workflow of both personnel and equipment is a problem currently faced and solved by many enterprises.

The Internet of things technology is introduced into the management of personnel and equipment in multiple industries, the conditions of the personnel and the equipment can be monitored in real time, the states of the personnel and the equipment can be known at any time, and the fault condition can be judged according to the states. Therefore, personnel and equipment can be effectively called, the supervision of the personnel is effectively monitored, and meanwhile, the equipment is fully utilized, so that various work is guaranteed to be smoothly carried out. The technology of the internet of things is introduced into the management and maintenance of various industries, and the management and maintenance work is pushed to a new era of information management by combining the internet technology.

Therefore, various detection algorithms are needed, and the real-time physical sign information of the personnel can be acquired by connecting the sensor with the mobile phone end; the positioning system and the sensor are connected with the mobile phone end, so that the working time and the working process of personnel can be obtained in real time, the working level of the personnel can be calculated through the obtained total working time, and the cost benefit of the personnel can be calculated; the mobile phone end is connected with the sensing system and the positioning system through control, the working time and the working state of the equipment can be positioned in real time, the equipment information is collected, the use times and the use rate of the equipment can be obtained, and the use energy consumption of the equipment can be effectively reduced through the opening and closing states of the remote control equipment.

In many industries at present, the monitoring of the working state of staff is mainly to position the staff, and technically, a single Bluetooth technology or an ultra-wideband technology is mainly adopted. The Bluetooth technology is adopted, so that the price is low, the power consumption is low, but the communication distance is short, and the anti-blocking capability is weak; the ultra-wideband technology is adopted, the price is high, the power consumption is large, the communication distance is slightly higher than that of the Bluetooth technology, but the positioning precision is poor, and the anti-blocking capability is weak. A unified device abstraction model is designed in the prior invention patent application document 'a massive Internet of things device access and management method' with the publication number of CN 113132192A; the device abstraction model is used for describing the capability characteristics of the device, and describing the capability of the device and the provided service through the abstraction model; and describing the device model in the form of JSON and XML, and uploading the model to a cloud end system and a device end after the device model is developed. The capability features of the device are device model identification attributes: device type, vendor, model, protocol type, services supported by the device, attributes, and control commands. The method has simple steps, and carries out standardized processing on massive heterogeneous equipment by designing an equipment unified description model. The technical scheme disclosed in the prior patent application document emphasizes on identifying equipment information, the equipment cannot be remotely controlled, the technical scheme cannot monitor and manage personnel while managing the equipment, the technical problem of waste of equipment energy consumption under application scenes that personnel are dynamically managed and are not on duty and the like cannot be solved, and the technical scheme in the prior application document does not realize the functions of personnel and enterprise management, assessment and the like, and cannot avoid the adverse effect of manual management. The utility model discloses a device and system of device and system that the utility model is CN214704680U intelligent asset location, state monitoring and management's device and system for intelligent asset location, state monitoring and management including being used for the wireless beacon of scanning discernment preset range in order to acquire asset positional information for detect asset operating condition's state detecting element, be used for thing networking communication unit and the main control unit with outside high in the clouds server communication, the utility model discloses a noninductive monitoring of valuable asset when unusual off-line is reported to the police, prevents that it from losing and stolen. The movement track recording of the asset equipment is realized by applying the low-power Bluetooth technology. The prior patent document does not disclose specific implementation of personnel management and assessment, and the technical scheme in the prior document cannot perform fine detection on personnel and equipment states simultaneously, cannot perform remote real-time control on abnormal information and special conditions, and has a single application scene in the prior art.

In most industries, the monitoring and control of the equipment are still in a manual mode, the mode needs a person to manage and maintain the equipment on duty, and the person on duty needs to regularly check whether the equipment is normally used, so that the equipment manager is inconvenient to know the service condition of the equipment in real time and monitor and manage the equipment in a unified manner, the operation and maintenance cost is high, and the utilization rate of equipment resources is low.

Therefore, the current personnel positioning scheme and the monitoring scheme for the equipment cannot meet the positioning and management requirements of many industries on the personnel and the equipment, and when enterprises need to expand the production, the internet of things for the personnel and the equipment becomes a problem which needs to be solved urgently in many industries at present in order to reduce unnecessary cost overhead, reduce internal consumption and improve customer satisfaction.

In conclusion, the prior art has the technical problems of dependence on manpower, higher management cost, lower management efficiency and lower operation precision.

Disclosure of Invention

The invention aims to solve the technical problems of dependence on labor, high management cost, low management efficiency and low operation precision in the prior art.

The invention adopts the following technical scheme to solve the technical problems: based on thing networking multi-scenario manager action and equipment operating mode action system includes:

a data acquisition module, comprising: the system comprises an intelligent terminal, wireless charging equipment and a communication system;

the intelligent terminal is used for monitoring, acquiring and transmitting personnel positioning control behavior data and equipment positioning control behavior data and is connected with the communication system;

the intelligent terminal includes: the device comprises a positioning module, a sensor module, an equipment sensing module, a power supply module and an abnormality detection processing module;

the positioning module is used for carrying out real-time positioning by combining Beidou satellite positioning and Bluetooth 4.0 to obtain real-time positioning information;

the sensor module includes: the physical sign detection module and the gyroscope are connected, and the sensor module is connected with the positioning module;

sign detection module for real-time supervision obtains personnel's sign information, wherein sign information includes: blood pressure, heart rate, body temperature and oxygen saturation index;

the gyroscope is used for judging the body posture of the person;

the equipment perception module for the perception acquires equipment operating condition time data, and equipment perception module is connected with sensor module, and equipment perception module includes: the system comprises an integrated controller and a control sensing node;

the integrated controller is respectively connected with the intelligent management platform and the control sensing node;

the control sensing nodes are installed in each tested room or equipment;

the abnormality detection processing module includes: a processor and a memory for detecting device anomalies, wherein the processor comprises:

the data set construction unit is used for constructing abnormal behavior data sets of personnel and equipment and randomly dividing the abnormal behavior data sets into a training set and a testing set according to a preset proportion;

the neural network construction unit is used for constructing a space-time separation 3D convolutional layer and a 3D space pyramid pooling layer to obtain a 3D convolutional neural network;

the supervised learning construction unit is used for constructing the classification of supervised learning to obtain a supervised learning model;

the neural network training unit is used for inputting a training set into the 3D convolutional neural network to obtain a trained 3D convolutional neural network model through continuous iterative training, and is connected with the data set construction unit and the neural network construction unit;

the supervised learning training unit is used for inputting the training set into the supervised learning model so as to realize sample classification, and is connected with the data set construction unit;

the abnormal behavior identification unit is used for monitoring and acquiring personnel positioning and entering a video stream in real time, inputting the video stream into a trained 3D convolutional neural network model, inputting the data of the voltage fluctuation condition of equipment into a supervised learning model, and identifying the abnormal behavior of the video monitoring and equipment use condition, wherein the abnormal behavior identification unit is connected with the neural network training unit and the supervised learning training unit;

the communication system comprises a preset NB-IoT communication module, a smart management platform and a communication module, wherein the preset NB-IoT communication module is used for sending the personnel positioning control behavior data and the equipment positioning control behavior data to the smart management platform, presenting the distribution situation of different scenes in a global manner, and sending the real-time positioning information, the personnel sign information and the body posture to the smart management platform through the preset NB-IoT communication module;

the personnel management module is used for acquiring basic information of the user through management operation and is connected with the intelligent terminal;

the device management module is used for acquiring device use information through management operation and is connected with the intelligent terminal;

the intelligent management platform is arranged at a mobile phone end of a supervisor, is used for receiving the physical sign information and the body posture of a person, is used for data storage and staff client analysis, draws a dynamic physical sign graph to obtain body condition information, gives out early warning when the abnormality of the person to be tested is judged, intelligently manages the staff or equipment according to the personnel positioning control behavior data and the equipment positioning control behavior data to obtain the working time of the staff and the equipment, processes the working time to obtain the staff working monitoring data and the equipment operation monitoring data, generates a report form according to statistics, and is connected with a communication system.

The invention provides a system for managing assets and behaviors of personnel and equipment, which can be applied to multiple scenes.

The abnormity detection processing module comprises a processor for detecting the abnormity of the equipment and a memory, wherein the memory is used for storing various collected data and computer programs, and the processor is used for calling and executing the computer programs from the memory, so that the intelligent terminal can acquire and collect the abnormal behaviors of the working scene of the personnel.

The invention gets rid of the dependence of manpower, connects the mobile phone end through the positioner, the heat sensing probe and the control sensing equipment, transmits data in real time, finely detects the states of personnel and equipment, remotely controls abnormal information and special conditions in real time, can be applied to monitoring of a plurality of scenes, realizes the systematization and data management between people and objects, and reduces the loss of manpower resources and equipment.

According to the invention, the sensor is connected with the mobile phone terminal, so that the physical sign information of the personnel can be acquired in real time, and the privacy of the personnel is protected.

In a more specific technical solution, the data set construction unit constructs the abnormal behavior data set by using the following logic:

D _T ＝α*D ₁ +β*D ₂ +γ*D ₃

wherein D is _T For abnormal behavior data sets, D ₁ To simulate an abnormal behavior data set, D ₂ For historical abnormal behavior data sets, D ₃ For the related abnormal behavior data set, alpha, beta and gamma are respectively the simulated abnormal behavior data set D ₁ Historical abnormal behavior data set D ₂ Related abnormal behavior data set D ₃ In abnormal behavior data set D _T And the proportion coefficient is alpha + beta + gamma is 1, and the abnormal behavior data set is randomly divided into a training set and a testing set according to a preset proportion.

The method also divides the training set and the test set of the model according to the proportionality coefficient when dividing the training set and the test set of the model, and can ensure the robustness of the model.

In a more specific technical scheme, a neural network construction unit adopts a network structure similar to a traditional convolutional neural network CNN, divides a 3 × 3 × 3 convolution kernel into 1 × 3 × 1 and 3 × 1 × 1 convolution kernels, and constructs a space-time separation 3D convolutional layer; adding an attention mechanism to the space-time separation 3D convolution layer, and constructing a space-time separation 3D convolution layer with the attention mechanism; adding a spatial pyramid pooling layer between the space-time separation 3D convolutional layer and the classifier, processing feature graphs with different lengths output by the 3D convolutional layer and outputting feature vectors with the same length; and providing the feature vectors output by the spatial pyramid pooling layer as fixed dimension vectors to a full-connection layer for classification to obtain a trained 3D convolutional neural network model.

The invention increases the attention mechanism to identify the area needing attention and increase the weight of the area; the 3D convolution is separated into the space direction convolution and the time direction convolution, so that the model fitting degree can be accelerated, and end-to-end training is facilitated.

The method adds an SPP network layer (a space pyramid pooling layer) between the attention mechanism-added space-time separation 3D convolutional layer and the classifier, processes according to the time dimension, performs maximum pooling operation, changes the time dimension into 19 dimensions after processing, can process feature graphs with different lengths output by the convolutional layer to output feature vectors with the same length, is suitable for processing videos with different lengths, and solves the problem that the traditional 3DCNN neural network requires that input data have the same dimension and behavior video blocks have different lengths.

In a more specific technical solution, the abnormal behavior identification unit processes the video stream with the following logic to identify the video monitoring abnormal behavior:

the three formulas are the changes of the characteristic diagram after 3D convolution calculation, wherein D _out To output the number of channels, D _in For input of channel number, H _out To output height, H _in For inputting height, W _out To output width, W _in For input width, padding is the padding size, default value for partition is 1, and kernel _ size is the size of the convolution kernel.

According to the invention, the real-time states of the personnel and the equipment are connected through the Internet, and the data are sent to the mobile phone end, so that the user can remotely monitor and control, the traditional manual monitoring mode is abandoned, and informatization and equipment management are organically combined, so that the equipment management not only achieves the effect of getting twice the result with half the effort, but also the overall management level is greatly improved.

In a more specific technical solution, the intelligent management platform further includes:

the personnel and equipment dynamic display component is used for inquiring and displaying the dynamic and distribution conditions of personnel and equipment in real time;

the personnel and equipment inquiry and scheduling component is used for inquiring the actual positions and the activity tracks of personnel and equipment so as to schedule management personnel and equipment;

and the personnel performance processing component records the arrival time, the departure time and the total working time of the personnel, and analyzes and judges the working time of the personnel according to the positioning data so as to evaluate the performance of the management personnel.

According to the invention, the positioning system and the sensor are connected with the mobile phone terminal, so that the working range, the working time and the working process of the staff can be acquired in real time, and the working content and the working level of the staff can be calculated through the total working time and the client reservation condition acquired by the mobile phone terminal. The dependence of traditional staff management on manpower is reduced, the staff management cost is reduced, and meanwhile, the management efficiency and the operation precision are improved.

the safety management module is used for setting the service time of the equipment and the temperature of the sensor, outputting equipment alarm information when the equipment fails, and outputting personnel body alarm information when the physical condition of a worker is detected to be abnormal;

the real-time monitoring module is used for searching the basic information and the position information of the staff or the equipment;

the enterprise assessment module is used for obtaining enterprise assessment scores according to the working conditions of workers, case handling conditions and safety management conditions;

and the enterprise management module is used for counting the personnel information, the total working time and the working content of the working personnel and forming a statistical report.

In a more specific technical scheme, the real-time supervision module processes personnel information acquired by the data acquisition module and combines output information of the personnel management module to count and present the safety state, the working time and the working place of a worker.

The invention perfects the statistics of the information of the working personnel and avoids the condition of operation omission; the real-time supervision module is arranged, so that the working condition can be fed back in time; the enterprise personnel safety effective management is facilitated, meanwhile, the enterprise bad behaviors and the total scores can be accurately checked, and the loss caused by the material reduction due to the labor stealing of the personnel is reduced.

In the existing enterprise management, all management is related to subjective behaviors of people, standardization and accurate quantification are rarely achieved, when enterprises need to expand reproduction, if the enterprises only depend on the experience of management personnel, the expansion speed is limited, and the human management is not efficient, so that the enterprise expansion reproduction is benefited after all management behaviors are detected and mechanized, and the human action is reduced as long as the application scale of equipment is expanded.

In a more specific technical solution, the supervised learning construction unit measures similarity between samples by calculating a distance, i.e. KNN algorithm, which includes: the distance of each sample point in the training set and test set is calculated with the following logic:

the following formula:

calculating to obtain an Euclidean distance;

the following formula:

calculating to obtain a single data point and a Mahalanobis distance between the data points, wherein Sigma is a covariance matrix of a multi-dimensional random variable, and mu is a sample mean value;

and sorting all the distance values, selecting the first k samples with the minimum distance, determining the frequency of the classification information in the k samples, and returning the classification with the highest frequency in the first k samples to obtain the final classification category.

The k value of the method can be obtained through cross validation and other modes, a strong consistency result can be obtained through the KNN algorithm, and the error rate of the KNN algorithm is reduced for some good k values along with the trend of infinite data.

In a more specific technical scheme, the enterprise management module is used for counting the use information of the equipment and forming a report, wherein the use information of the equipment comprises the use times, the running time and the use rate of the equipment.

According to the invention, the control sensing system and the positioning system are connected with the mobile phone end, so that the working time and the working state of the equipment can be positioned in real time, the equipment can be remotely controlled, and the obtained equipment information is summarized, thereby effectively reducing the energy consumption of the equipment.

The system and the method have the advantages that in enterprise management, the capability level of personnel and the working efficiency of equipment play a vital role in the satisfaction degree of customers, and the system and the method can effectively monitor and manage the personnel and the equipment, so that the system and the method can be suitable for multiple scenes, such as application scenes of families, medical and beauty, hospitals, construction sites, factory production lines, internal parking lots, emergency epidemic prevention management and the like, and can effectively solve the management problems of large personnel mobility, high capital circulation, unnecessary equipment energy consumption and the like aiming at the enterprises.

In a more specific technical scheme, the method for managing the behavior of the personnel and the behavior of the equipment working condition based on the multi-scenario of the Internet of things comprises the following steps:

s1, monitoring, acquiring and transmitting the personnel positioning control behavior data and the equipment positioning control behavior data, wherein the step S1 comprises the following steps:

s101, carrying out real-time positioning by combining Beidou satellite positioning and Bluetooth 4.0 to obtain real-time positioning information;

s102, monitoring and acquiring physical sign information of the personnel in real time, wherein the physical sign information comprises: blood pressure, heart rate, body temperature and oxygen saturation index;

s103, judging the body posture of the person;

s2, sensing and acquiring the working state time data of the equipment;

s3, constructing abnormal behavior data sets of personnel and equipment, and randomly dividing the abnormal behavior data sets into a training set and a testing set according to a preset proportion;

s4, constructing a space-time separation 3D convolution layer and a 3D space pyramid pooling layer to obtain a 3D convolution neural network;

s5, constructing classification of supervised learning so as to obtain a supervised learning model;

and S6, inputting the training set into the constructed 3D convolutional neural network for continuous iterative training. Obtaining a trained 3D convolutional neural network model;

s7, inputting the training set into the constructed supervised learning model to realize sample classification;

s8, monitoring and acquiring personnel positioning and entering video stream in real time, inputting the video stream into the trained 3D convolutional neural network model, inputting the data of the voltage fluctuation condition of the equipment into a supervised learning model, and identifying the abnormal behaviors of the video monitoring and the equipment use condition;

s9, constructing a space-time separation 3D convolution layer and a 3D space pyramid pooling layer for increasing attention mechanism to obtain a 3D convolution neural network;

s10, inputting the training set into the constructed 3D convolutional neural network for continuous iterative training to obtain a trained 3D convolutional neural network model;

s11, monitoring and acquiring personnel positioning and equipment voltage fluctuation conditions in real time, forming data into a video stream, inputting the video stream into a trained 3D convolutional neural network model, processing the video stream, and recognizing video monitoring abnormal behaviors;

s12, sending the personnel positioning control behavior data and the equipment positioning control behavior data to an intelligent management platform to present the distribution situation of the different scenes in a global manner, and sending the real-time positioning information, the personnel sign information and the body posture to the intelligent management platform through a preset NB-IoT communication module;

s13, acquiring user basic information through management operation;

s14, acquiring equipment use information through management operation;

and S15, receiving the personnel sign information and the body posture, performing data storage and staff client analysis, drawing a dynamic sign graph to obtain body condition information, giving out an early warning when the detected personnel is judged to be abnormal, performing intelligent management on the personnel or equipment according to the personnel positioning control behavior data and the equipment positioning control behavior data to obtain the working time of the personnel and the equipment, processing to obtain staff working monitoring data and equipment running monitoring data, and generating a report according to statistics.

Compared with the prior art, the invention has the following advantages: the invention provides a system for managing assets and behaviors of personnel and equipment, which can be applied to multiple scenes.

According to the invention, the positioning system and the sensor are connected with the mobile phone end, so that the working range, the working time and the working process of the staff can be obtained in real time, and the working content and the working level of the staff can be calculated through the total working time and the client reservation condition obtained by the mobile phone end. The dependence of traditional staff management on manpower is reduced, the staff management cost is reduced, and meanwhile, the management efficiency and the operation precision are improved.

In the existing enterprise management, all management is related to subjective behaviors of people, standardization and accurate quantification can be realized rarely, when enterprises need to expand reproduction, if the enterprises only depend on the experience of management personnel, the expansion speed is limited, and the human management is not efficient, so that the enterprise expansion reproduction is benefited after the invention detects and machines all management behaviors, and the human action is reduced as long as the application scale of equipment is expanded.

The system and the method have the advantages that in enterprise management, the capability level of personnel and the working efficiency of equipment play a vital role in the satisfaction degree of customers, and the system and the method can effectively monitor and manage the personnel and the equipment, so that the system and the method can be suitable for multiple scenes, such as application scenes of families, medical and beauty, hospitals, construction sites, factory production lines, internal parking lots, emergency epidemic prevention management and the like, and can effectively solve the management problems of large personnel mobility, high capital circulation, unnecessary equipment energy consumption and the like aiming at the enterprises. The invention solves the technical problems of dependence on manpower, higher management cost, lower management efficiency and lower operation precision in the prior art.

Drawings

Fig. 1 is a schematic diagram of a data acquisition module of an internet-of-things multi-scenario manager behavior and equipment condition behavior system according to embodiment 1 of the present invention;

FIG. 2 is a schematic diagram of a processor unit of an exception detection processing module according to embodiment 1 of the present invention;

FIG. 3 is a schematic diagram of the components of a personnel management module according to embodiment 1 of the present invention;

FIG. 4 is a schematic diagram of a person monitoring procedure in embodiment 1 of the present invention;

FIG. 5 is a schematic view of the monitoring step of the apparatus according to embodiment 1 of the present invention;

FIG. 6 is a schematic view of an anomaly detection process according to embodiment 1 of the present invention;

FIG. 7 is a diagram illustrating specific steps of data set construction according to embodiment 1 of the present invention;

fig. 8 is a schematic flow chart of a 3D convolutional neural network construction according to embodiment 1 of the present invention;

fig. 9 is a schematic diagram illustrating the intelligent management platform module according to embodiment 1 of the present invention;

fig. 10 is a schematic structural diagram of a terminal of an internet of things scene acquisition device in embodiment 2 of the present invention;

fig. 11 is a schematic structural diagram of an internet of things acquisition device in embodiment 2 of the present invention;

fig. 12 is a diagram of a prototype terminal of an internet of things signal acquisition device according to embodiment 2 of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the embodiments of the present invention, and it is obvious that the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

Example 1

The invention discloses a multi-scenario manager behavior and equipment working condition behavior system based on the Internet of things, which comprises: data acquisition module 1, personnel management module 2, equipment management module 3 and wisdom management platform 4.

As shown in fig. 1, the data acquisition module 1 includes an intelligent terminal 11, a wireless charging device 12 for charging the intelligent terminal, and a communication system 13. The intelligent terminal 11 monitors the positioning and control behaviors of personnel and equipment, and transmits data to the supervision platform in real time through the communication system 13. The communication system 13 includes an NB-IoT module 131, and data transmission between the operator base station and the monitoring platform is implemented by NB-IoT communication. In the embodiment, by adopting a new generation of communication technology NB-IoT communication technology and an Internet of things technology, real-time positioning of personnel and equipment is realized, the distribution situation in the whole scene is globally displayed, the movement track of any personnel or equipment can be inquired, the monitoring is more comprehensive, the management and control are stronger, and the efficiency is higher; the method has the advantages of ultra-low power consumption, mass connection, high coverage rate and ultra-low cost.

The intelligent terminal 11 comprises a positioning module 111, a sensor module 112, an equipment sensing module 113, a power supply module 114 and an abnormality detection processing module 115; the positioning module 111 performs real-time positioning by combining Beidou satellite positioning and Bluetooth 4.0, and sends real-time positioning information to the intelligent management platform 4 through the NB-IoT communication module; the sensor module 12 includes a sign detection module 121 and a gyroscope 122, the sign detection module 121 monitors sign information of a person in real time, and the sign information includes blood pressure, heart rate, body temperature and an oxygen saturation index; the gyroscope 122 determines the body posture of the person; the sensor module 12 sends the measured information to the intelligent management platform 4 through the NB-IoT communication module; the power module 114 supplies power to other modules; the device sensing module 113 is used for sensing the working state and working time of the device.

The device sensing module 113 includes an integrated controller 1131 and a control sensing node 1132, the integrated controller 1131 is connected to the smart management platform 4 and the control sensing node 1132, the control sensing node 1132 is installed in each room or device to be tested, the integrated controller 1131 includes an information reading module 11311, a cyclic query module 11312 and a command sending module 11313, the control sensing node 1132 includes a monitoring module 11321 and a device control module 11322, the information reading module 11311 is configured to receive operation information sent by the platform and then parse a configuration information file to obtain address information of each control sensing node 1132 and a control command and a device type of each device to be controlled, the cyclic query module 11312 queries and stores monitoring data of the corresponding control sensing node 1132 according to the obtained address information, the command sending module 11313 is configured to send the obtained device type and control command to the corresponding control sensing node 1132, the monitoring module 11321 is configured to monitor a working state of the device where the control sensing node 1132 is located, and the device control module 11322 is configured to control the device to be tested according to the received device type and the control command.

The anomaly detection processing module 115 comprises a processor 1151 and a memory 1152 for detecting device anomalies, the memory 1152 is used for storing collected various data and computer programs, the processor 1151 is used for calling and executing the computer programs from the memory 1152, so that the intelligent terminal 11 can obtain and collect anomaly behaviors of a personnel working scene, and when the anomaly detection processing module 115 detects anomalies, the intelligent terminal 11 sends anomaly information to the intelligent management platform 4 through the NB-IoT communication module 131.

As shown in fig. 2, in the abnormality detection processing module 115, the processor 1151 for identifying and detecting the behavior of the human work scenario abnormality includes:

the data set construction unit 11511 is configured to construct abnormal behavior data sets of personnel and equipment, and randomly divide the abnormal behavior data sets into a training set and a test set according to a preset proportion;

a neural network construction unit 11512 configured to construct a space-time separation 3D convolutional layer and a 3D spatial pyramid pooling layer, so as to obtain a 3D convolutional neural network;

a supervised learning construction unit 11513 configured to construct a supervised learning classification to obtain a supervised learning model;

the neural network training unit 11514 is configured to input a training set to the constructed 3D convolutional neural network for continuous iterative training to obtain a trained 3D convolutional neural network model, and the neural network training unit 11514 is connected with the data set construction unit 11511 and the neural network construction unit 11512;

the supervised learning training unit 11515 is configured to input a training set to the constructed supervised learning model to realize sample classification, and the supervised learning training unit 11515 is connected with the data set construction unit 11511;

the abnormal behavior recognition unit 11516 is configured to monitor and acquire personnel positioning entering video streams in real time, input the video streams into the trained 3D convolutional neural network model, input data of the equipment voltage fluctuation condition into the supervised learning model, and complete abnormal behavior recognition of video monitoring and equipment use conditions, and the abnormal behavior recognition unit 11516 is connected with the neural network training unit 11514 and the supervised learning training unit 11515.

The intelligent management platform 4 is installed at a mobile phone end of a supervisor, receives sensor data sent by the intelligent terminal 11, stores and analyzes vital sign information of staff and customers, draws a dynamic sign graph to obtain body condition information of a measured target, and gives an early warning when the body condition of the measured staff is abnormal; the intelligent management platform 4 intelligently manages the staff or the equipment according to the positioning data sent by the intelligent terminal 11, receives the working time of the staff through positioning, calculates the working efficiency through the average reception client time, sets the working time of the equipment, and generates a report form by counting the equipment utilization rate, the equipment utilization times, the staff workload and the staff cost and income.

The physical sign information comprises blood pressure, heart rate, body temperature and oxygen saturation index.

Positioning module 111 adopts big dipper satellite positioning module 1111 to combine with bluetooth module 1112, carries out real-time location, and wherein, bluetooth module 1112 can adopt for example bluetooth 4.0:

aiming at the situation that when a worker works outdoors, the Beidou satellite automatically reads the position information of the worker, and transmits the position information to the supervision platform for positioning; aiming at the personnel needing to work at the fixed post, the Bluetooth reading equipment pre-installed at the position of the fixed post automatically reads the personnel information, and the supervision platform judges the position of the personnel according to the read position of the Bluetooth equipment; aiming at the personnel working in the room, the gridding management is carried out on each layer of the indoor building, each grid is provided with corresponding Bluetooth reading equipment, once the personnel enters the corresponding grid, the equipment automatically reads the information of the personnel, and the supervision platform judges the position of the personnel according to the read position of the Bluetooth equipment.

The management platform sends the positioning data to carry out intelligent management to staff and equipment according to intelligent terminal 11 specifically:

inquiring and displaying the dynamic and distribution conditions of personnel and equipment at a certain place at any time point;

inquiring the actual position and the activity track of the personnel or the equipment so as to carry out reasonable scheduling management on the personnel or the equipment;

recording the arrival time, the departure time and the total working time of related personnel at any place, and analyzing and judging the working time of the personnel according to the positioning data so as to conveniently evaluate and manage the performance of the personnel;

the data acquisition module 1 is used for carrying out data analysis according to the data acquired by the intelligent terminal 11; the data acquisition module 1 comprises personnel data acquisition and equipment data acquisition, wherein the personnel data acquisition is used for inputting personnel information, the equipment data acquisition is used for inputting equipment information, and the personnel information is the working time, the working content and the working flow of personnel; the equipment information is the working time and the working items of the equipment;

as shown in fig. 3, the personnel management module 2 is used for basic information management of a user, and includes a personal information management module 21, a others information management module 22 and a function management module 23, the personal information management module 21 is used for providing password modification and information modification functions, the others information management module 22 is used for giving management authority to others, the function management module 21 is used for implementing a client's reservation operation, wherein the personal information management module 21 further includes a password modification module 211 and an information modification module 212, the others information management module 22 includes an authority giving module 221, and the function management module 23 further includes a reservation module 231;

referring to fig. 4, in the present embodiment, the specific steps of monitoring the personnel include:

s1, acquiring the position information and the physical sign information of the person to be measured through the sensing equipment and the positioning equipment;

s2, transmitting the information to an intelligent management platform through communication equipment after the information is collected;

s3, judging whether the personnel in the fixed post area are on duty or not;

s4, if yes, judging whether the personnel sign information is normal;

s5, if not, sending alarm information to the intelligent management platform;

and S6, if the physical sign information of the personnel is normal, counting the workload of the personnel according to the working time of the personnel on duty, forming a report and sending the report to the enterprise management module.

The device management module 3 is used for managing the use information of the device, and the use information comprises the use times and the use rate of the device;

the equipment management module 3 senses the working state and working time of the equipment through the intelligent terminal, and the positioning system inquires the current position of the employee; sensing a basic working state through sensor equipment under the condition of not invading privacy; the abnormal detection processor is used for detecting whether the equipment has abnormal behaviors or not, so that management digitization is achieved.

As shown in fig. 5, in the present embodiment, the monitoring procedure for the device includes:

s1', obtaining the position information and working state of the tested equipment through the positioning equipment and the control sensing equipment;

s2', acquiring the on-duty state of the staff in the detected area through the positioning equipment;

s3', transmitting the information to the intelligent management platform through the communication equipment after the information is collected;

s4', judging whether there is employee working on duty;

s5', if not, automatically starting the ventilation and disinfection equipment, and closing all other equipment;

s6', if yes, judging whether the equipment processes the working state;

s7', counting the equipment utilization rate and the use times according to the use time of the equipment to be tested, forming a report and sending the report to the enterprise management module;

s8', according to the device use information and motion trail of other areas obtained by the positioning device, the device is set to standby state, so that the device can be scheduled flexibly.

As shown in fig. 6, the behavior recognition and detection for the abnormality of the human work scenario in the abnormality detection processing module includes: in this embodiment, the process of identifying and detecting the behavior of the abnormality in the working scene of the person includes the following steps:

s101, constructing abnormal behavior data sets of personnel and equipment, and randomly dividing the abnormal behavior data sets into a training set and a testing set according to a preset proportion;

s102, constructing a space-time separation 3D convolutional layer and a 3D space pyramid pooling layer to obtain a 3D convolutional neural network;

s103, building classification of supervised learning to obtain a supervised learning model;

s104, inputting the training set into the constructed 3D convolutional neural network for continuous iterative training to obtain a trained 3D convolutional neural network model;

s105, inputting the training set into the constructed supervised learning model to realize sample classification;

and S106, monitoring and acquiring personnel positioning and entering a video stream in real time, inputting the video stream into the trained 3D convolutional neural network model, inputting the data of the voltage fluctuation condition of the equipment into the supervised learning model, and completing video monitoring and abnormal behavior identification of the service condition of the equipment.

As shown in fig. 7, in this embodiment, the step S101 further includes the following specific steps:

s1011, predefining a plurality of types of abnormal behaviors, respectively simulating the plurality of types of abnormal behaviors, carrying out video acquisition, and constructing a simulated abnormal behavior data set;

s1012, video collection is carried out on the abnormal behavior cases which occur historically, and a historical abnormal behavior data set is constructed;

s1013, video acquisition is carried out on the related abnormal behavior cases inquired on the network, and a related abnormal behavior data set is constructed;

s1014, performing data collection on the simulation abnormal behavior data set, the historical abnormal behavior data set and the related abnormal behavior data set according to a preset proportionality coefficient to construct an abnormal behavior data set;

and S1015, randomly dividing the abnormal behavior data set into a training set and a testing set according to a preset proportion and preset proportion coefficients of the three data sets.

In the aspect of data set construction, a plurality of abnormal behaviors can be predefined according to specific application scenes, the defined abnormal behaviors are simulated respectively, and a simulated abnormal behavior data set D is constructed by collecting the simulated abnormal behaviors ₁ . Meanwhile, historical abnormal behavior cases occurring in a real environment and related abnormal behavior cases inquired on the internet are combined to be collected to respectively construct a historical abnormal behavior data set D ₂ And associated abnormal behavior data set D ₃ Combining the data sets according to a preset proportionality coefficient to construct an abnormal behavior data set D _T ：

D _T ＝α*D ₁ +β*D ₂ +γ*D ₃ Wherein D is _T For abnormal behavior data sets, D ₁ To simulate an abnormal behavior data set, D ₂ For historical abnormal behavior data sets, D ₃ For the related abnormal behavior data set, alpha, beta and gamma are respectively the simulated abnormal behavior data set D ₁ Historical abnormal behavior data set D ₂ Related abnormal behavior data set D ₃ In abnormal behavior data set D _T And α + β + γ is 1. When the model training set and the test set are divided, the model training set and the test set are also divided according to the proportionality coefficient, so that the robustness of the model can be ensured.

As shown in fig. 8, in this embodiment, the step S102 further includes the following specific steps:

s1021, adopting a network structure similar to a traditional convolutional neural network CNN, dividing a 3 × 3 × 3 convolution kernel into 1 × 3 × 1 and 3 × 1 × 1 convolution kernels, and constructing a space-time separation 3D convolutional layer;

s1022, adding a spatial pyramid pooling layer between the space-time separation 3D convolutional layer and the classifier, processing feature graphs with different lengths output by the 3D convolutional layer and outputting feature vectors with the same length;

and S1023, providing the feature vectors output by the spatial pyramid pooling layer as fixed dimension vectors to a full-connection layer for classification to obtain a trained 3D convolutional neural network model.

The space pyramid pooling layer is added between the space-time separation 3D convolutional layer and the classifier, the space pyramid pooling layer is processed in a time dimension, the largest pooling operation is performed, the time dimension meets requirements after processing is completed, feature graphs with different lengths are output by the convolutional layer, feature vectors with the same length are output, the method is suitable for processing videos with different lengths, the problem that the size of a traditional CNN input image needs to be fixed is solved, and the aspect ratio and the size of the input image can be arbitrary. And finally, providing the feature vectors output by the spatial pyramid pooling layer as fixed dimension vectors for a full-connection layer for classification, so that end-to-end training can be realized, and the hyper-parameters can be adjusted to obtain a trained 3D convolutional neural network model.

In addition, in this embodiment, in step S104, the following logic may be adopted for inputting the video stream to the 3D convolutional neural network model for performing the 3D convolutional operation output dimension calculation:

the three formulas are the changes of the characteristic diagram after 3D convolution calculation, D _out To output the number of channels, D _in For input of channel number, H _out To output height, H _in For inputting height, W _out To output width, W _in For input width, padding is the padding size, default value for partition is 1, and kernel _ size is the size of the convolution kernel.

In this embodiment, the foregoing step S103 further includes: the similarity between the samples is measured by calculating the distance, namely a KNN algorithm, and the method comprises the following steps:

calculating the distance of each sample point in the training set and the test set, sequencing all distance values, selecting the first k samples with the minimum distance, determining the frequency of classification information in the k samples, returning the classification with the highest frequency in the first k samples, and obtaining the final classification.

Common distance measures include Euclidean distance, Mahalanobis distance, etc., and the Euclidean distance formula is

The mahalanobis distance formula for a single data point is:

the mahalanobis distance between data points x, y is formulated as:

where Σ is the covariance matrix of the multidimensional random variable, and μ is the sample mean.

The k value can be obtained through cross validation and other modes, a strong consistency result can be obtained through the KNN algorithm, and the error rate of the KNN algorithm is reduced for some good k values along with the tendency of infinite data.

The method for obtaining the supervised learning model further comprises the following steps: KMeans, DBScan, etc. algorithms;

in this embodiment, in step S104, inputting the training set to the constructed 3D convolutional neural network for continuous iterative training to obtain a trained 3D convolutional neural network model, further including:

and inputting the test set into the constructed 3D convolutional neural network for continuous verification, and performing gradient return and parameter updating on the network parameters by using a back propagation algorithm.

In this embodiment, the step of identifying and detecting the abnormal behavior of the human work scene further includes:

when the abnormal behavior is identified in the video monitoring, the information such as the key frame, the abnormal behavior description, the abnormal behavior occurrence time and the like of the abnormal behavior video is recorded, and the data is sent to the safety management module in the intelligent management platform.

In this embodiment, the method for identifying and detecting abnormal behavior of a human work scene further includes:

As shown in fig. 9, the intelligent management platform 4 is configured to receive data information uploaded by the device of the matching staff, and make a report of the obtained information.

The intelligent management platform 4 comprises a safety management module 41, a real-time supervision module 42, an enterprise assessment module 43 and an enterprise management module 44;

the intelligent management platform 4 comprises a safety management module 41, which is used for setting the service time of the equipment, fixing the working time of the equipment so as to reduce resource waste and outputting alarm information when the equipment fails; the safety management module is also used for setting the temperature of the sensor and outputting alarm information when the physical condition of the detected object is detected and the physical condition of the detected person is abnormal;

the intelligent management platform 4 further comprises a real-time supervision module 42; the real-time monitoring module 42 is used for accurately searching the basic information and the position information of the staff or the equipment;

the real-time supervision module 42 is used for processing the personnel information acquired by the data acquisition module, then counting the safety state, the working time and the working place of the workers according to the output information of the personnel management module, and displaying the safety state, the working time and the working place of the workers to the real-time supervision module 42;

the real-time monitoring module 42 is further configured to count the device information acquired by the data acquisition module, combine the output information of the device management module, count the working state, the working time, and the number of times of use of the device to be tested, and display the count to the real-time monitoring module 42;

the intelligent management platform 4 further comprises an enterprise assessment module 43; the enterprise assessment module 43 is used for obtaining enterprise assessment scores according to the working conditions of workers, case handling conditions and safety management conditions;

the intelligent management platform 4 further comprises an enterprise management module 44; the enterprise management module 44 is used for counting the personnel information, the total working time and the working content of the workers and forming a statistical report;

the enterprise management module 44 is further configured to count and form a report on the usage information of the device, where the usage information of the device includes the number of times of usage of the device, the running time, and the usage rate of the device.

Example 2

The implementation scene of the invention is applied to hospitals as an example, and comprises a data acquisition module 1, a personnel management module 2, an equipment management module 3 and an intelligent management platform 4;

for example, after a doctor in a certain hospital enters a work post, the position and the personnel sign information of the doctor are monitored in real time through the positioning device and the sensor device in the data acquisition module 1, in the embodiment, the internet of things acquisition device can adopt, for example, an internet of things scene acquisition device terminal and send the information to the intelligent management platform 4, each module in the intelligent management platform 4 correspondingly counts the working time of the internet of things, and the function module in the personnel management module 2 can count the client reservation condition and count the working content and the working level of the doctor by combining the working time of the client reservation condition;

the enterprise assessment module 43 acquires the position information of the staff by combining with the positioning equipment in the data acquisition module, finds the corresponding staff through the staff management module, and obtains the information of the working range, the working content, the working process and the like of the staff, so as to obtain the working condition and the attendance qualification rate of the staff;

the enterprise management module 44 combines the data of other modules, and is used for obtaining enterprise assessment scores according to the working conditions of workers, case handling conditions and safety management conditions and forming a report, and based on the scores, enterprises can calculate wages, bonus and the like according to regulations;

the physical sign information of the personnel can be monitored in real time through the sensor equipment in the data acquisition module 1, the information is sent to the intelligent management platform 4, the physical sign information of the personnel can be monitored in real time through the safety management module 41 and the real-time supervision module 42 in the intelligent management platform 4, the privacy of customers is protected, meanwhile, the epidemic situation is effectively prevented and controlled, and if the physical sign information is abnormal, the intelligent management platform 4 can give an alarm;

whether personnel in the area are on duty or not can be monitored through the positioning equipment and the control sensing equipment in the data acquisition module 1, and when the personnel are not monitored in the area, the control sensing equipment automatically starts disinfection and ventilation equipment in the area;

the equipment is positioned in real time and data are sent to the intelligent management platform 4 through positioning equipment and control sensing equipment in the data acquisition module 1 and a real-time supervision module 42 of the equipment management module 3 and the intelligent management platform 4, and information such as the working time and working items of the equipment is acquired, so that the use times and the equipment use rate of the equipment can be counted;

the enterprise management module 44 combines the device data acquired by the data acquisition module, and obtains the energy consumption of the device according to the information such as the service time and the number of times of use of the device, thereby calculating the remaining life of the device and forming a report.

As shown in fig. 10, the terminal of the scene collection device of the internet of things includes a forcible entry prevention contact 1 ', a 5GPCB antenna 2 ' and an AI control main board 3 '.

As shown in fig. 11 and 12, the internet of things collecting device includes an input general control power supply 4 ', an anti-dismantling outer frame 5 ', a power bent socket 6 ', a power straight socket 7 ' and an AI control main board 8 '.

In conclusion, the invention provides a system which can be applied to multiple scenes and is used for managing assets and behaviors of personnel and equipment, on one hand, the system is simple in structure, reasonable in design, simple to operate, strong in practicability and excellent in control performance, on the other hand, the service life of the equipment can be prolonged, the working efficiency of the personnel is improved, and the aim of reducing the cost is fulfilled.

The method also divides the model training set and the test set according to the proportionality coefficient when dividing the model training set and the test set, and can ensure the robustness of the model.

The attention increasing mechanism can identify the area needing attention and increase the weight of the area; the 3D convolution is separated into the space direction convolution and the time direction convolution, so that the model fitting degree can be accelerated, and end-to-end training is facilitated.

The above examples are only intended to illustrate the technical solution of the present invention, and not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims

1. Based on thing networking multi-scenario manager action and equipment operating mode action system, its characterized in that, the system includes:

the intelligent terminal comprises: the device comprises a positioning module, a sensor module, an equipment sensing module, a power supply module and an abnormality detection processing module;

the sign detection module is used for monitoring and acquiring the sign information of the personnel in real time, wherein the sign information comprises: blood pressure, heart rate, body temperature and oxygen saturation index;

the gyroscope is used for judging the body posture of a person;

the equipment perception module is used for perceiving and acquiring time data of the working state of the equipment, the equipment perception module is connected with the sensor module, and the equipment perception module comprises: the system comprises an integrated controller and a control sensing node;

the control sensing nodes are installed in each tested room or equipment;

the abnormality detection processing module includes: a processor and a memory to detect device exceptions, wherein the processor comprises:

the neural network training unit is used for inputting the training set to the 3D convolutional neural network so as to continuously perform iterative training to obtain a trained 3D convolutional neural network model, and is connected with the data set construction unit and the neural network construction unit;

the abnormal behavior identification unit is used for monitoring and acquiring personnel positioning entering video stream in real time, inputting the video stream into the trained 3D convolutional neural network model, inputting data of equipment voltage fluctuation conditions into the supervised learning model so as to identify abnormal behaviors of video monitoring and equipment use conditions, and is connected with the neural network training unit and the supervised learning training unit;

the communication system comprises a preset NB-IoT communication module, is used for sending the personnel positioning control behavior data and the equipment positioning control behavior data to an intelligent management platform, presents different scene distribution conditions in a global manner, and sends the real-time positioning information, the personnel physical sign information and the body posture to the intelligent management platform through the preset NB-IoT communication module, and is connected with the intelligent terminal;

the equipment management module is used for acquiring equipment use information through management operation and is connected with the intelligent terminal;

the intelligent management platform is installed at a mobile phone end of a supervisor and used for receiving the personnel sign information and the body posture, storing data and analyzing staff clients, drawing a dynamic sign graph to obtain body condition information, sending out early warning when the detected staff is judged to be abnormal, intelligently managing staff or equipment according to the personnel positioning control behavior data and the equipment positioning control behavior data to obtain the working time of the staff and the equipment, processing the working time to obtain staff working monitoring data and equipment running monitoring data, and generating a report according to statistics, wherein the intelligent management platform is connected with the communication system.

2. The IOT (Internet of things) -based multi-scenario manager behavior and equipment working condition behavior system according to claim 1, wherein the data set construction unit constructs an abnormal behavior data set by using the following logic:

D _T ＝α*D ₁ +β*D ₂ +γ*D ₃

wherein D is _T For abnormal behavior data sets, D ₁ To simulate an abnormal behavior data set, D ₂ For historical abnormal behavior data sets, D ₃ For the related abnormal behavior data set, alpha, beta and gamma are respectively the simulated abnormal behavior data set D ₁ Historical abnormal behavior data set D ₂ Related abnormal behavior data set D ₃ In abnormal behavior data set D _T And the abnormal behavior data set is randomly divided into a training set and a testing set according to a preset proportion, wherein the proportion coefficient is alpha + beta + gamma is 1.

3. The internet of things multi-scenario manager behavior and equipment working condition behavior system based on claim 1, wherein the neural network training unit is configured to divide a 3 × 3 × 3 convolution kernel into 1 × 3 × 1 and 3 × 1 × 1 convolution kernels by using a network structure similar to a traditional Convolution Neural Network (CNN) to construct a space-time separation 3D convolution layer; adding an attention mechanism to the space-time separation 3D convolutional layer, and constructing a space-time separation 3D convolutional layer with the attention mechanism; adding a space pyramid pooling layer between the space-time separation 3D convolution layer and the classifier, processing feature graphs with different lengths output by the 3D convolution layer and outputting feature vectors with the same length; and providing the feature vectors output by the spatial pyramid pooling layer as fixed dimension vectors to a full-connection layer for classification to obtain a trained 3D convolutional neural network model.

4. The IOT (Internet of things) -based multi-scenario manager behavior and equipment working condition behavior system according to claim 1, wherein the abnormal behavior recognition unit processes the video stream with the following logic to recognize video monitoring abnormal behavior:

the three formulas are the changes of the characteristic diagram after 3D convolution calculation, wherein D _out To output the number of channels, D _in Is the input channel number, H _out To output height, H _in For inputting height, W _out To output width, W _in Is input intoWidth, padding is the padding size, partition is default value of 1, and kernel _ size is the size of the convolution kernel.

5. The internet of things multi-scenario manager behavior and device behavior based system according to claim 1, wherein the intelligent management platform further comprises:

the personnel and equipment inquiry and scheduling component is used for inquiring the actual positions and the activity tracks of the personnel and the equipment so as to schedule and manage the personnel and the equipment;

and the personnel performance processing component records the arrival time, the departure time and the total working time of the personnel, analyzes and judges the working time of the personnel according to the positioning data and evaluates and manages the performance of the personnel.

6. The internet of things multi-scenario manager behavior and device behavior based system according to claim 1, wherein the intelligent management platform further comprises:

the enterprise assessment module is used for obtaining enterprise assessment scores according to the working conditions of the workers, the case handling conditions and the safety management conditions;

and the enterprise management module is used for counting the personnel information, the total working time and the working content of the workers and forming a statistical report.

7. The IOT (Internet of things) -based multi-scenario manager behavior and equipment working condition behavior system according to claim 6, wherein the real-time supervision module processes personnel information collected by the data collection module and combines output information of the personnel management module to count and present the safety state, working time and working place of the personnel.

8. The Internet of things multi-scenario manager behavior and equipment condition behavior system based on claim 6, wherein the supervised learning construction unit measures similarity between samples by calculating distance, namely a KNN algorithm, which includes: the distance of each sample point in the training set and test set is calculated with the following logic:

the following formula:

calculating to obtain an Euclidean distance;

the following formula:

sorting all the distance values, selecting the first k samples with the minimum distance, determining the frequency of the classification information in the k samples, returning the classification with the highest frequency in the first k samples, and obtaining the final classification category.

9. The internet of things multi-scenario manager behavior and equipment working condition behavior system based on claim 6, wherein the enterprise management module is used for counting the use information of the equipment and forming a report, wherein the use information of the equipment comprises the use times, the running time and the equipment use rate of the equipment.

10. The method for managing the behavior of personnel and the behavior of equipment working conditions based on multiple scenes of the Internet of things is characterized by comprising the following steps:

s1, monitoring, acquiring and transmitting personnel positioning control behavior data and equipment positioning control behavior data, wherein the step S1 comprises the following steps:

s102, monitoring and acquiring physical sign information of personnel in real time, wherein the physical sign information comprises: blood pressure, heart rate, body temperature and oxygen saturation index;

s103, judging the body posture of the person;

s2, sensing and acquiring the working state time data of the equipment;

s5, establishing classification of supervised learning so as to obtain a supervised learning model;

and S6, inputting the training set into the constructed 3D convolutional neural network for continuous iterative training. Obtaining the trained 3D convolutional neural network model;

s8, monitoring and acquiring personnel positioning and entering video stream in real time, inputting the video stream into the trained 3D convolutional neural network model, and inputting the data of the equipment voltage fluctuation condition into the supervised learning model so as to identify the abnormal behaviors of video monitoring and equipment use conditions;

s11, monitoring and acquiring personnel positioning and equipment voltage fluctuation conditions in real time, forming data into a video stream, inputting the video stream into the trained 3D convolutional neural network model, processing the video stream, and recognizing video monitoring abnormal behaviors;

s12, sending the personnel positioning control behavior data and the equipment positioning control behavior data to an intelligent management platform to present different scene distribution conditions in a global manner, and sending the real-time positioning information, the personnel physical sign information and the body posture to the intelligent management platform through the preset NB-IoT communication module;

s13, acquiring user basic information through management operation;

s14, acquiring equipment use information through management operation;

and S15, receiving the personnel sign information and the body posture, performing data storage and staff customer analysis, drawing a dynamic sign graph to obtain body condition information, giving out early warning when the detected personnel is judged to be abnormal, performing intelligent management on the personnel or equipment according to the personnel positioning control behavior data and the equipment positioning control behavior data to obtain the working time of the personnel and the equipment, processing to obtain staff working monitoring data and equipment operation monitoring data, and generating a report according to statistics.