An intelligent inspection method based on the fusion of the Internet of Things and virtual reality based on AI big model

Abstract

The present invention relates to the technical field of intelligent inspection, and discloses an intelligent inspection method based on the fusion of the Internet of Things and virtual reality based on an AI large model, the method comprising: constructing a power grid anomaly recognition model to perform anomaly recognition on pre-processed sequence data of the Internet of Things device; constructing a power grid anomaly positioning model according to the anomaly recognition results of the power grid area where the Internet of Things device is located, and optimizing and solving to obtain abnormal position information in the power grid area; and driving the inspection robot to perform intelligent inspection based on virtual reality technology according to the abnormal position information. The present invention obtains the spatiotemporal characteristics of the power grid area at the power consumption level according to the power consumption information of the Internet of Things device, and calculates the spatiotemporal information of power consumption that characterizes the stability of power consumption in the power grid area to realize power grid anomaly recognition, and drives the inspection robot to perform intelligent inspection based on virtual reality technology according to the abnormal position information of the power grid area, so as to realize multiple power grid line inspections based on virtual reality and rapid generation of maintenance processing steps.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent inspection, and in particular to an intelligent inspection method based on the fusion of Internet of Things and virtual reality based on an AI big model. Background Art

[0002] In the electric power industry, the safe and stable operation of the power grid is crucial to the development of the social economy. However, traditional power grid inspection methods usually consume a lot of manpower and material resources, and there are certain safety hazards. At the same time, the power grid is widely distributed and has a large number of devices, making it difficult for traditional inspection methods to achieve comprehensive coverage and timely detection of potential problems. In response to this problem, the present invention proposes an intelligent inspection method based on the fusion of the Internet of Things and virtual reality based on an AI large model. By combining the perception and data transmission technology of the Internet of Things and the simulation and visualization technology of virtual reality, the efficiency of power grid inspection can be improved, costs can be reduced, and potential safety hazards can be effectively reduced. Summary of the invention

[0003] In view of this, the present invention provides an intelligent inspection method based on the fusion of Internet of Things and virtual reality based on AI big model, the purpose is to: 1) collect current, voltage and coordinate data of Internet of Things devices in the power grid area, based on the spatial distance between the coordinates of the Internet of Things device and the power grid equipment and the timing characteristic value under the preset timing step, use the AI ​​big model to obtain the spatiotemporal characteristics of the power grid area at the power consumption level, calculate the power consumption spatiotemporal information of the power grid area according to the spatiotemporal characteristics of the power grid area and the correlation of spatiotemporal characteristics, and obtain abnormal recognition results of different power grid areas, wherein the larger the spatiotemporal characteristics, the greater the current and voltage timing fluctuations of the Internet of Things devices in the same power grid area, and the higher the power consumption spatiotemporal information of the power grid area, the greater the current and voltage timing fluctuations of most Internet of Things devices in the power grid area, so as to realize power grid abnormality identification based on Internet of Things devices Normal recognition; 2) According to the abnormal recognition results of the power grid area where the IoT device is located, the power grid topology structure of the power grid area with the abnormality is simulated, and a reward function is set to iterate the abnormal position information, and the inspection robot is driven to perform intelligent inspection based on virtual reality technology according to the abnormal position information. The inspection robot transmits the acquired video image to the virtual reality device, and the virtual reality device performs three-dimensional virtual modeling on the video image. The operator wears the virtual reality device to conduct further safety inspection of the three-dimensional virtual scene, and uses the operating device to perform power grid line maintenance processing in the virtual scene on the power grid lines in the three-dimensional virtual scene, and sends the power grid line maintenance processing steps to the maintenance station closest to the power grid line location for line maintenance, thereby realizing multiple power grid line inspections based on virtual reality and rapid generation of maintenance processing steps.

[0004] To achieve the above purpose, the present invention provides an intelligent inspection method based on the fusion of the Internet of Things and virtual reality based on an AI big model, comprising the following steps:

[0005] S1: Using IoT sensors to collect IoT device sequence data, where the IoT device sequence data includes current, voltage and coordinate data of IoT devices, and transmitting the collected IoT device sequence data to a backend server for preprocessing to obtain preprocessed IoT device sequence data;

[0006] S2: Constructing a power grid anomaly recognition model to perform anomaly recognition on the preprocessed IoT device sequence data, wherein the power grid anomaly recognition model takes the preprocessed IoT device sequence data as input and takes the anomaly recognition result of the power grid area where the IoT device is located as output, wherein spatiotemporal correlation analysis is the main implementation method of the power grid anomaly recognition model;

[0007] S3: construct a power grid anomaly location model based on the anomaly identification results of the power grid area where the IoT device is located, and optimize and solve it to obtain the anomaly location information in the power grid area. The power grid anomaly location model simulates the power grid topology information in a digital way, and locates the anomaly in combination with the anomaly identification results of the power grid area. The Markov decision process is the main implementation method of the power grid anomaly location model.

[0008] S4: According to the abnormal position information, the inspection robot is driven to perform intelligent inspection based on virtual reality technology to obtain the power grid inspection results.

[0009] As a further improvement method of the present invention:

[0010] Optionally, in step S1, collecting IoT device sequence data using IoT sensors includes:

[0011] The IoT sensor is used to collect IoT device serial data, where the IoT device serial data includes the current, voltage and coordinate data of the IoT device. The collected IoT device serial data is:

[0012] {z n =(I n (t),U n (t),(x n ,y n ))|n∈[1,N]}

[0013] I n (t)=(I n (t1),I n (t2),...,I n (t m ),...,I n (t M ))

[0014] U n(t)=(U n (t1),U n (t2),...,U n (t m ),...,U n (t M ))

[0015] in:

[0016] z n Represents the current, voltage and coordinate data of the nth IoT device, I n (t) represents the current data of the nth IoT device, U n (t) represents the voltage data of the nth IoT device, t represents the timing information, (x n ,y n ) represents the plane coordinates of the nth IoT device, x n Indicates the plane coordinate value of the nth IoT device in the horizontal direction, y n Indicates the plane coordinate value of the nth IoT device in the vertical direction;

[0017] I n (t m ) indicates that the nth IoT device senses the time t at the mth sensor m Current value, t M represents the current sensor sensing moment, and the time difference between adjacent sensor sensing moments is Δt;

[0018] U n (t m ) indicates that the nth IoT device senses the time t at the mth sensor m Voltage value;

[0019] The collected IoT device serial data is transmitted to the backend server for preprocessing to obtain the preprocessed IoT device serial data {Z n =(I′ n (t),U′ n (t),(x n ,y n ))|n∈[1,N]},H n Represents the preprocessed current, voltage and coordinate data of the nth IoT device, I′ n (t) represents the preprocessed current data of the nth IoT device, U′ n (t) represents the preprocessed voltage data of the nth IoT device.

[0020] Optionally, the transmitting the collected IoT device sequence data to a background server for preprocessing includes:

[0021] The collected IoT device serial data is transmitted to the backend server for preprocessing, where the current, voltage and coordinate data z of the nth IoT device in the IoT device serial data are n The preprocessing process is:

[0022] S11: Opposite n Current data I n (t) Perform preprocessing to obtain the preprocessed current data I′ of the nth IoT device n (t):

[0023] I′ n (t) = (I′ n (t1),I′ n (t2),...,I′ n (t m ),...,I′ n (t M ))

[0024]

[0025] in:

[0026] I max Indicates the preset maximum current, I min Indicates the preset minimum current value;

[0027] I′ n (t m ) represents the current value I n (t m )’s preprocessing results;

[0028] S12: Opposite n Voltage data U n (t) Perform preprocessing to obtain the preprocessed voltage data I′ of the nth IoT device n (t):

[0029] U′ n (t) = (U′ n (t1),U′ n (t2),...,U′ n (t m ),...,U′ n (t M ))

[0030]

[0031] in:

[0032] U max Indicates the preset maximum voltage, Umin Indicates the preset minimum voltage value;

[0033] U′ n (t m ) represents the voltage value U n (t m )’s preprocessing results;

[0034] S13: Constructs the pre-processed current, voltage and coordinate data of the nth IoT device:

[0035] Z n =(I′ n (t),U′ n (t),(x n ,y n ))

[0036] in:

[0037] Z n Represents the preprocessed current, voltage, and coordinate data of the nth IoT device.

[0038] Optionally, constructing a power grid anomaly identification model in step S2 includes:

[0039] Construct a power grid anomaly recognition model, and use the power grid anomaly recognition model to perform anomaly recognition on the preprocessed IoT device sequence data. The power grid anomaly recognition model takes the preprocessed IoT device sequence data as input and takes the anomaly recognition result of the power grid area where the IoT device is located as output, wherein spatiotemporal correlation analysis is the main implementation method of the power grid anomaly recognition model, and the power grid anomaly recognition model includes an input layer, a spatiotemporal feature mapping layer, a spatiotemporal feature correlation calculation layer, and anomaly recognition layer;

[0040] The input layer is used to divide the preprocessed IoT device sequence data according to the power grid area, and obtain the preprocessed current, voltage and coordinate data of all IoT devices in each power grid area;

[0041] The spatiotemporal feature mapping layer is used to calculate the spatiotemporal features of each power grid area;

[0042] The temporal and spatial feature correlation calculation layer is used to calculate the temporal and spatial feature correlation of each power grid area;

[0043] The anomaly recognition layer is used to obtain anomaly recognition results for different power grid areas based on the spatiotemporal characteristics of the power grid area and the correlation of spatiotemporal characteristics.

[0044] Optionally, the using the power grid anomaly recognition model to perform anomaly recognition on the preprocessed IoT device sequence data includes:

[0045] The power grid anomaly recognition model is used to identify anomalies in the preprocessed IoT device sequence data, where the anomaly recognition process is as follows:

[0046] S21: The input layer receives the preprocessed IoT device sequence data and divides the preprocessed IoT device sequence data according to the power grid area to obtain the preprocessed current, voltage and coordinate data of all IoT devices in each power grid area, where the power grid area Ω k The pre-processed current, voltage and coordinate data of all IoT devices are:

[0047]

[0048] in:

[0049] Represents the power grid area Ω k Preprocessed current data of the sth IoT device Voltage data and coordinate data S k Represents the power grid area Ω k The total number of IoT devices in the area; in the embodiment of the present invention, each power grid area includes a number of boost stations, step-down stations and distribution stations;

[0050] k∈[1,K], K represents the total number of power grid areas, Ω1,Ω2,...,Ω K represents K power grid areas;

[0051] S22: The spatiotemporal feature mapping layer calculates the spatiotemporal features of each power grid area, where the power grid area Ω k The spatiotemporal characteristics are:

[0052]

[0053]

[0054] in:

[0055] W represents the feature mapping matrix, and T represents transpose;

[0056] F k Represents the power grid area Ω k The spatial and temporal characteristics of

[0057] Represents the power grid area Ω k The spatiotemporal information of power consumption of the s-th IoT device;

[0058] exp(·) represents an exponential function with a natural constant as the base. Represents coordinate data Arrival at the power grid areaΩk The shortest distance of the power grid equipment, including the step-up substation, step-down substation and distribution substation in the power grid area;

[0059] Represents the power grid area Ω k The current timing characteristic value of the sth IoT device in, Represents the power grid area Ω k The voltage timing characteristic value of the s-th IoT device in;

[0060] The power grid area Ω k The calculation process of the current timing characteristic value and voltage timing characteristic value of the sth IoT device is as follows:

[0061] S221: For the power grid area Ω k The preprocessed current and voltage data of the s-th IoT device are reconstructed, and the reconstruction result is:

[0062]

[0063]

[0064]

[0065]

[0066] in:

[0067] L represents a preset timing step; in the embodiment of the present invention, L is set to 2;

[0068] Represents the current data after preprocessing The current reconstruction result is represents the reconstruction component in the current reconstruction result, Represents the power grid area Ω k The sth IoT device in t h The pre-processed current value at the moment;

[0069] Represents the voltage data after preprocessing The voltage reconstruction result is represents the reconstruction component in the voltage reconstruction result, Represents the power grid area Ω k The sth IoT device in t h The pre-processed voltage value at time;

[0070] S222: Perform time series distance screening on the reconstruction components in the current reconstruction result and the voltage reconstruction result respectively:

[0071] Under the condition that the time step is L, for the reconstructed component Filter and reconstruct components from the current reconstruction results The Euclidean distance between them is lower than the preset current distance threshold The reconstruction components of the in Represents the current data after preprocessing (t) standard deviation;

[0072] Under the condition that the time step is L, for the reconstructed component Filter and reconstruct components from the voltage reconstruction results The Euclidean distance between them is lower than the preset voltage distance threshold The reconstruction components of the in Represents the voltage data after preprocessing The standard deviation of

[0073] S223: Calculate the power grid area Ω k The current timing characteristic value and voltage timing characteristic value of the sth IoT device in are:

[0074]

[0075]

[0076]

[0077]

[0078] in:

[0079] They are the current information and voltage information under the timing step L respectively;

[0080] S23: The spatiotemporal feature correlation calculation layer calculates the spatiotemporal feature correlation of each power grid area, where the power grid area Ω k The spatial and temporal feature correlation of is:

[0081]

[0082]

[0083] in:

[0084] Per k Represents the power grid area Ω k The spatial and temporal characteristics of the correlation;

[0085] Represents the power grid area Ωk The spatial and temporal feature correlation of the s-th IoT device in;

[0086] ||·|| represents the L1 norm;

[0087] S24: The anomaly recognition layer obtains anomaly recognition results of different power grid areas according to the spatiotemporal characteristics and the correlation of spatiotemporal characteristics of the power grid area, where the power grid area Ω k The abnormal recognition results are:

[0088]

[0089] in:

[0090] ρ k Represents the power grid area Ω k The abnormal recognition result of k =1 indicates the power grid area Ω k There is an abnormality, ρ k =0 means the power grid area Ω k There are no anomalies;

[0091] Represents the power grid area Ω k The spatiotemporal information of electricity consumption of the sth IoT device in If the L2 norm of k The sth IoT device is marked as an IoT device with abnormal spatiotemporal information of electricity consumption;

[0092] ε represents the preset abnormal threshold.

[0093] Optionally, in step S3, a power grid anomaly location model is constructed and optimized to obtain abnormal location information in the power grid area, including:

[0094] According to the abnormal identification results of the power grid area where the IoT device is located, the power grid area with abnormalities is selected to construct the corresponding power grid abnormality positioning model, and the constructed power grid abnormality positioning model is optimized and solved to obtain the abnormal location information in the power grid area. The power grid abnormality positioning model simulates the power grid topology information in a digital way, and combines the abnormal identification results of the power grid area to perform abnormal positioning. The Markov decision process is the main implementation method of the power grid abnormality positioning model, in which the abnormal power grid area The abnormal location information determination process is as follows:

[0095] S31: Digitally monitor the grid area The grid topology information is simulated to obtain the grid area The topological structure diagram G=(E,V), where E represents the power grid area The location set of IoT devices in , V represents the power grid area The set of power grid lines between any two equipment locations in;

[0096] S32: Get the power grid area The power consumption spatiotemporal information of all IoT devices in the grid is collected, and the IoT devices are clustered based on the power consumption spatiotemporal information to obtain the power grid area The Q IoT device clusters in the network are calculated, the location mean of all IoT devices in each IoT device cluster is calculated, and the location mean is mapped in the power grid area. The grid line location with the closest mean value in the middle distance is taken as the grid area The Q initial abnormal position information in:

[0097] (Loc0(1),Loc0(2),...,Loc0(q),...,Loc0(Q))

[0098] in:

[0099] Loc0(q) represents the grid area Ω k The location mean mapping result of the qth IoT device cluster in *, that is, the power grid area The qth initial abnormal position information in ;

[0100] S33: Set the current iteration number of the initial abnormal location information to d, the initial value of d is 0, and the maximum value is Max, then the power grid area The d-th iteration result of the q-th initial abnormal location information in is Loc d (q), constructs the reward function of the initial abnormal position information iteration based on the Markov decision process, where the initial abnormal position information is obtained from Loc d (q) Iterate to Loc d The reward function for (q+1) is:

[0101] Reward(Loc d (q+1))=avg(Loc d (q+1))―avg(Loc d (q))

[0102] in:

[0103] Reward(Loc d (q+1)) indicates that the initial abnormal location information is from Loc d (q) Iterate to Loc d The reward function of (q+1);

[0104] avg(Loc d (q)) indicates the location Locd (q) The average distance along the power grid line to the IoT device with abnormal spatiotemporal information of power consumption in the qth IoT device cluster;

[0105] S34: Iterate the Q initial abnormal position information based on the reward function, select the power line position that makes the reward function positive as the d+1th iteration result of the Q initial abnormal position information, until the power line position cannot be selected, and use the Q positions iterated at this time as the power grid area Abnormal location information.

[0106] Optionally, in step S4, driving the inspection robot to perform intelligent inspection based on virtual reality technology according to the abnormal position information includes:

[0107] According to the abnormal position information, the inspection robot is driven to perform intelligent inspection based on virtual reality technology to obtain the power grid inspection results, where the power grid inspection results include the location of the damaged power grid line and the corresponding power grid line maintenance and processing steps. The intelligent inspection process is as follows:

[0108] S41: The inspection robot receives the position in the abnormal position information and arrives at the corresponding power grid line position;

[0109] S42: The inspection robot turns on a camera to obtain a video image of the power grid line location it has reached;

[0110] S43: the inspection robot transmits the acquired video image to the virtual reality device, and the virtual reality device performs three-dimensional virtual modeling on the video image to obtain a three-dimensional virtual scene of the power grid line location; in the embodiment of the present invention, the virtual reality device includes VR glasses and a VR helmet;

[0111] S44: comparing the three-dimensional virtual scene of the power grid line location with the three-dimensional scene of the normal power grid line. If there is a difference, it means that the power grid line location is damaged, and the operator wears the virtual reality device to further check the three-dimensional virtual scene, and generates an operating device in the three-dimensional virtual scene. In the embodiment of the present invention, the operator can stretch and perform various transformation operations on the three-dimensional virtual scene of the power grid line location to conduct further safety inspection;

[0112] S45: The operator uses the operating device to perform grid line maintenance processing on the grid line in the three-dimensional virtual scene in the virtual scene, and sends the grid line maintenance processing steps to the maintenance station closest to the grid line location. The maintenance station performs grid line maintenance according to the grid line maintenance processing steps.

[0113] In order to solve the above problem, the present invention provides an electronic device, the electronic device comprising:

[0114] A memory storing at least one instruction;

[0115] Communication interface, enabling electronic equipment to communicate; and

[0116] The processor executes the instructions stored in the memory to implement the above-mentioned intelligent inspection method based on the integration of the Internet of Things and virtual reality based on the AI ​​big model.

[0117] In order to solve the above problems, the present invention also provides a computer-readable storage medium, in which at least one instruction is stored. The at least one instruction is executed by a processor in an electronic device to implement the above-mentioned intelligent inspection method based on the integration of the Internet of Things and virtual reality based on the AI ​​big model.

[0118] Compared with the prior art, the present invention proposes an intelligent inspection method based on the fusion of the Internet of Things and virtual reality based on an AI big model, which has the following advantages:

[0119] First, this scheme proposes a power grid anomaly identification method, which uses the power grid anomaly identification model to identify anomalies in the preprocessed IoT device sequence data. The anomaly identification process is as follows: the input layer receives the preprocessed IoT device sequence data, and divides the preprocessed IoT device sequence data according to the power grid area, and obtains the preprocessed current, voltage and coordinate data of all IoT devices in each power grid area, where the power grid area Ω k The pre-processed current, voltage and coordinate data of all IoT devices are:

[0120]

[0121] in: Represents the power grid area Ω k Preprocessed current data of the sth IoT device Voltage data and coordinate data S k Represents the power grid area Ω k The total number of IoT devices in the grid; k∈[1,K], K represents the total number of grid areas, Ω1,Ω2,...,Ω K represents K power grid areas; the spatiotemporal feature mapping layer calculates the spatiotemporal features of each power grid area, where the power grid area Ω k The spatiotemporal characteristics are:

[0122]

[0123]

[0124] Where: W represents the feature mapping matrix, T represents transposition; Fk Represents the power grid area Ω k The spatial and temporal characteristics of Represents the power grid area Ω k The spatiotemporal information of electricity consumption of the s-th IoT device in ; exp(·) represents an exponential function with a natural constant as the base, Represents coordinate data Arrival at the power grid areaΩ k The shortest distance of the power grid equipment, including the step-up substation, step-down substation and distribution substation in the power grid area; Represents the power grid area Ω k The current timing characteristic value of the sth IoT device in, Represents the power grid area Ω k The voltage time series characteristic value of the sth IoT device in the network; the time-space characteristic correlation calculation layer calculates the time-space characteristic correlation of each power grid area, where the power grid area Ω k The spatial and temporal feature correlation of is:

[0125]

[0126]

[0127] Where: Per k Represents the power grid area Ω k The spatial and temporal characteristics of the correlation; Represents the power grid area Ω k The spatial-temporal feature correlation of the sth IoT device in ; ||·|| represents the L1 norm; the anomaly recognition layer obtains the anomaly recognition results of different power grid areas according to the spatial-temporal features and spatial-temporal feature correlation of the power grid area, where the power grid area Ω k The abnormal recognition results are:

[0128]

[0129] Where: k Represents the power grid area Ω k The abnormal recognition result of k =1 indicates the power grid area Ω k There is an abnormality, ρ k =0 means the power grid area Ω k There are no anomalies; Represents the power grid area Ω k The spatiotemporal information of electricity consumption of the sth IoT device in If the L2 norm of kThe sth IoT device is marked as an IoT device with abnormal spatiotemporal information of electricity consumption; ε represents the preset abnormal threshold. This scheme collects the current, voltage and coordinate data of IoT devices in the power grid area, obtains the spatiotemporal characteristics of the power grid area at the power consumption level based on the spatial distance between the coordinates of the IoT device and the power grid equipment and the time series characteristic value under the preset time series step, calculates the spatiotemporal information of electricity consumption in the power grid area according to the spatiotemporal characteristics of the power grid area and the correlation of spatiotemporal characteristics, and obtains the abnormal identification results of different power grid areas. The larger the spatiotemporal characteristics, the greater the time series fluctuation of the current and voltage of IoT devices in the same power grid area. The higher the spatiotemporal information of electricity consumption in the power grid area, the greater the current and voltage time series fluctuation of most IoT devices in the power grid area, and realizes the power grid abnormality identification based on IoT devices.

[0130] At the same time, this scheme proposes a method for locating abnormal position information of a power grid area. According to the abnormal identification results of the power grid area where the IoT device is located, the power grid area with abnormalities is selected to construct a corresponding power grid abnormality positioning model, and the constructed power grid abnormality positioning model is optimized and solved to obtain abnormal position information in the power grid area. The power grid abnormality positioning model simulates the power grid topology information in a digital way, and performs abnormal positioning in combination with the abnormal identification results of the power grid area. The abnormal location information determination process is as follows: digitally identify the power grid area The grid topology information is simulated to obtain the grid area The topological structure diagram G=(E,V), where E represents the power grid area The location set of IoT devices in , V represents the power grid area The set of power grid lines between any two device locations in; Get the power grid area The power consumption spatiotemporal information of all IoT devices in the grid is collected, and the IoT devices are clustered based on the power consumption spatiotemporal information to obtain the power grid area The Q IoT device clusters in the network are calculated, the location mean of all IoT devices in each IoT device cluster is calculated, and the location mean is mapped in the power grid area. The grid line location with the closest mean value in the middle distance is taken as the grid area The Q initial abnormal position information in:

[0131] (Loc0(1),Loc0(2),...,Loc0(q),...,Loc0(Q))

[0132] Where: Loc0(q) represents the power grid area The location mean mapping result of the qth IoT device cluster in the grid area The qth initial abnormal location information in the grid; the current iteration number of the initial abnormal location information is set to d, the initial value of d is 0, and the maximum value is Max, then the grid area The d-th iteration result of the q-th initial abnormal location information in is Loc d (q), constructs the reward function of the initial abnormal position information iteration based on the Markov decision process, where the initial abnormal position information is obtained from Loc d (q) Iterate to Loc d The reward function for (q+1) is:

[0133] Reward(Loc d (q+1))=avg(Loc d (q+1))―avg(Loc d (q))

[0134] Among them: Reward(Loc d (q+1)) indicates that the initial abnormal location information is from Loc d (q) Iterate to Loc d The reward function of (q+1); avg(Loc d (q)) indicates the location Loc d (q) The average distance along the power grid line to the IoT device with abnormal spatiotemporal information of power consumption in the qth IoT device cluster; iterate the Q initial abnormal position information based on the reward function, and select the power grid line position that makes the reward function positive as the d+1th iteration result of the Q initial abnormal position information, until the power grid line position cannot be selected, and the Q positions iterated at this time are used as the power grid area Abnormal location information. This solution simulates the power grid topology structure of the power grid area where the IoT device is located based on the abnormal identification results of the power grid area, and sets a reward function to iterate the abnormal location information. According to the abnormal location information, the inspection robot is driven to perform intelligent inspection based on virtual reality technology. The inspection robot transmits the acquired video image to the virtual reality device. The virtual reality device performs three-dimensional virtual modeling on the video image. The operator wears the virtual reality device to conduct further safety inspection of the three-dimensional virtual scene, and uses the operating device to perform power grid line maintenance processing in the virtual scene on the power grid lines in the three-dimensional virtual scene. The power grid line maintenance processing steps are sent to the maintenance station closest to the power grid line location for line maintenance, thereby realizing multiple power grid line inspections based on virtual reality and rapid generation of maintenance processing steps. BRIEF DESCRIPTION OF THE DRAWINGS

[0135] Figure 1A flowchart of an intelligent inspection method based on the fusion of the Internet of Things and virtual reality based on an AI big model provided by one embodiment of the present invention;

[0136] Figure 2 A schematic diagram of the structure of an electronic device for implementing an intelligent inspection method based on the fusion of the Internet of Things and virtual reality based on an AI big model provided by one embodiment of the present invention.

[0137] In the figure: 1 electronic device, 10 processor, 11 memory, 12 program, 13 communication interface.

[0138] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0139] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0140] The embodiment of the present application provides an intelligent inspection method based on the fusion of the Internet of Things and virtual reality based on an AI big model. The execution subject of the intelligent inspection method based on the fusion of the Internet of Things and virtual reality based on the AI ​​big model includes but is not limited to at least one of the electronic devices such as a server and a terminal that can be configured to execute the method provided by the embodiment of the present application. In other words, the intelligent inspection method based on the fusion of the Internet of Things and virtual reality based on the AI ​​big model can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc.

[0141] Embodiment 1:

[0142] S1: Use IoT sensors to collect IoT device serial data, where the IoT device serial data includes current, voltage and coordinate data of IoT devices, and transmit the collected IoT device serial data to a background server for preprocessing to obtain preprocessed IoT device serial data.

[0143] In the step S1, the IoT sensor is used to collect IoT device sequence data, including:

[0144] The IoT sensor is used to collect IoT device serial data, where the IoT device serial data includes the current, voltage and coordinate data of the IoT device. The collected IoT device serial data is:

[0145] {z n =(I n (t),U n (t),(x n ,y n))|n∈[1,N]}

[0146] I n (t)=(I n (t1),I n (t2),...,I n (t m ),...,I n (t M ))

[0147] U n (t)=(U n (t1),U n (t2),...,U n (t m ),...,U n (t M ))

[0148] in:

[0149] z n Represents the current, voltage and coordinate data of the nth IoT device, I n (t) represents the current data of the nth IoT device, U n (t) represents the voltage data of the nth IoT device, t represents the timing information, (x n ,y n ) represents the plane coordinates of the nth IoT device, x n Indicates the plane coordinate value of the nth IoT device in the horizontal direction, y n Indicates the plane coordinate value of the nth IoT device in the vertical direction;

[0150] I n (t m ) indicates that the nth IoT device senses the time t at the mth sensor m Current value, t M represents the current sensor sensing moment, and the time difference between adjacent sensor sensing moments is Δt;

[0151] U n (t m ) indicates that the nth IoT device senses the time t at the mth sensor m Voltage value;

[0152] The collected IoT device serial data is transmitted to the backend server for preprocessing to obtain the preprocessed IoT device serial data {Z n =(I′ n (t),U′ n (t),(x n ,y n))|n∈[1,N]},H n Represents the preprocessed current, voltage and coordinate data of the nth IoT device, I′ n (t) represents the preprocessed current data of the nth IoT device, U′ n (t) represents the preprocessed voltage data of the nth IoT device.

[0153] The method of transmitting the collected IoT device serial data to the backend server for preprocessing includes:

[0154] The collected IoT device serial data is transmitted to the backend server for preprocessing, where the current, voltage and coordinate data z of the nth IoT device in the IoT device serial data are n The preprocessing process is:

[0155] S11: Opposite n Current data I n (t) Perform preprocessing to obtain the preprocessed current data I′ of the nth IoT device n (t):

[0156] I′ n (t) = (I′ n (t1),I′ n (t2),...,I′ n (t m ),...,I′ n (t M ))

[0157]

[0158] in:

[0159] I max Indicates the preset maximum current, I min Indicates the preset minimum current value;

[0160] I′ n (t m ) represents the current value I n (t m )’s preprocessing results;

[0161] S12: Opposite n Voltage data U n (t) Perform preprocessing to obtain the preprocessed voltage data I′ of the nth IoT device n (t):

[0162] U′ n (t) = (U′ n (t1),U′ n(t2),...,U′ n (t m ),...,U′ n (t M ))

[0163]

[0164] in:

[0165] U max Indicates the preset maximum voltage, U min Indicates the preset minimum voltage value;

[0166] U′ n (t m ) represents the voltage value U n (t m )’s preprocessing results;

[0167] S13: Constructs the pre-processed current, voltage and coordinate data of the nth IoT device:

[0168] Z n =(I′ n (t),U′ n (t),(x n ,y n ))

[0169] in:

[0170] Z n Represents the preprocessed current, voltage, and coordinate data of the nth IoT device.

[0171] S2: Construct a power grid anomaly recognition model to perform anomaly recognition on the preprocessed IoT device sequence data. The power grid anomaly recognition model takes the preprocessed IoT device sequence data as input and takes the anomaly recognition result of the power grid area where the IoT device is located as output.

[0172] The step S2 constructs a power grid anomaly recognition model, including:

[0173] Construct a power grid anomaly recognition model, and use the power grid anomaly recognition model to perform anomaly recognition on the preprocessed IoT device sequence data. The power grid anomaly recognition model takes the preprocessed IoT device sequence data as input and takes the anomaly recognition result of the power grid area where the IoT device is located as output, wherein spatiotemporal correlation analysis is the main implementation method of the power grid anomaly recognition model, and the power grid anomaly recognition model includes an input layer, a spatiotemporal feature mapping layer, a spatiotemporal feature correlation calculation layer, and anomaly recognition layer;

[0174] The input layer is used to divide the preprocessed IoT device sequence data according to the power grid area, and obtain the preprocessed current, voltage and coordinate data of all IoT devices in each power grid area;

[0175] The spatiotemporal feature mapping layer is used to calculate the spatiotemporal features of each power grid area;

[0176] The temporal and spatial feature correlation calculation layer is used to calculate the temporal and spatial feature correlation of each power grid area;

[0177] The anomaly recognition layer is used to obtain anomaly recognition results for different power grid areas based on the spatiotemporal characteristics of the power grid area and the correlation of spatiotemporal characteristics.

[0178] The method of using the power grid anomaly recognition model to perform anomaly recognition on the preprocessed IoT device sequence data includes:

[0179] The power grid anomaly recognition model is used to identify anomalies in the preprocessed IoT device sequence data, where the anomaly recognition process is as follows:

[0180] S21: The input layer receives the preprocessed IoT device sequence data and divides the preprocessed IoT device sequence data according to the power grid area to obtain the preprocessed current, voltage and coordinate data of all IoT devices in each power grid area, where the power grid area Ω k The pre-processed current, voltage and coordinate data of all IoT devices are:

[0181]

[0182] in:

[0183] Represents the power grid area Ω k Preprocessed current data of the sth IoT device Voltage data and coordinate data S k Represents the power grid area Ω k The total number of IoT devices in the area; in the embodiment of the present invention, each power grid area includes a number of boost stations, step-down stations and distribution stations;

[0184] k∈[1,K], K represents the total number of power grid areas, Ω1,Ω2,...,Ω K represents K power grid areas;

[0185] S22: The spatiotemporal feature mapping layer calculates the spatiotemporal features of each power grid area, where the power grid area Ω k The spatiotemporal characteristics are:

[0186]

[0187]

[0188] in:

[0189] W represents the feature mapping matrix, and T represents transpose;

[0190] F k Represents the power grid area Ω k The spatial and temporal characteristics of

[0191] Represents the power grid area Ω k The spatiotemporal information of power consumption of the s-th IoT device;

[0192] exp(·) represents an exponential function with a natural constant as the base. Represents coordinate data Arrival at the power grid areaΩ k The shortest distance of the power grid equipment, including the step-up substation, step-down substation and distribution substation in the power grid area;

[0193] Represents the power grid area Ω k The current timing characteristic value of the sth IoT device in, Represents the power grid area Ω k The voltage timing characteristic value of the s-th IoT device in;

[0194] The power grid area Ω k The calculation process of the current timing characteristic value and voltage timing characteristic value of the sth IoT device is as follows:

[0195] S221: For the power grid area Ω k The preprocessed current and voltage data of the s-th IoT device are reconstructed, and the reconstruction result is:

[0196]

[0197]

[0198]

[0199]

[0200] in:

[0201] L represents a preset timing step; in the embodiment of the present invention, L is set to 2;

[0202] Represents the current data after preprocessing The current reconstruction result is represents the reconstruction component in the current reconstruction result, Represents the power grid area Ω k The sth IoT device in t h The pre-processed current value at the moment;

[0203] Represents the voltage data after preprocessing The voltage reconstruction result is represents the reconstruction component in the voltage reconstruction result, Represents the power grid area Ω k The sth IoT device in t h The pre-processed voltage value at time;

[0204] S222: Perform time series distance screening on the reconstruction components in the current reconstruction result and the voltage reconstruction result respectively:

[0205] Under the condition that the time step is L, for the reconstructed component Filter and reconstruct components from the current reconstruction results The Euclidean distance between them is lower than the preset current distance threshold The reconstruction components of the in Represents the current data after preprocessing (t) standard deviation;

[0206] Under the condition that the time step is L, for the reconstructed component Filter and reconstruct components from the voltage reconstruction results The Euclidean distance between them is lower than the preset voltage distance threshold The reconstruction components of the in Represents the voltage data after preprocessing The standard deviation of

[0207] S223: Calculate the power grid area Ω k The current timing characteristic value and voltage timing characteristic value of the sth IoT device in are:

[0208]

[0209]

[0210]

[0211]

[0212] in:

[0213] They are the current information and voltage information under the timing step L respectively;

[0214] S23: The spatiotemporal feature correlation calculation layer calculates the spatiotemporal feature correlation of each power grid area, where the power grid area Ω k The spatial and temporal feature correlation of is:

[0215]

[0216]

[0217] in:

[0218] Per k Represents the power grid area Ω k The spatial and temporal characteristics of the correlation;

[0219] Represents the power grid area Ω k The spatial and temporal feature correlation of the s-th IoT device in;

[0220] ||·|| represents the L1 norm;

[0221] S24: The anomaly recognition layer obtains anomaly recognition results of different power grid areas according to the spatiotemporal characteristics and the correlation of spatiotemporal characteristics of the power grid area, where the power grid area Ω k The abnormal recognition results are:

[0222]

[0223] in:

[0224] ρ k Represents the power grid area Ω k The abnormal recognition result of k =1 indicates the power grid area Ω k There is an abnormality, ρ k =0 means the power grid area Ω k There are no anomalies;

[0225] Represents the power grid area Ω k The spatiotemporal information of electricity consumption of the sth IoT device in If the L2 norm of k The sth IoT device is marked as an IoT device with abnormal spatiotemporal information of electricity consumption;

[0226] ε represents the preset abnormal threshold.

[0227] S3: A power grid anomaly location model is constructed based on the anomaly identification results of the power grid area where the IoT device is located, and an optimization solution is performed to obtain the abnormal location information in the power grid area. The power grid anomaly location model simulates the power grid topology information in a digital manner, and performs anomaly location in combination with the anomaly identification results of the power grid area.

[0228] In the step S3, a power grid anomaly location model is constructed and optimized to obtain abnormal location information in the power grid area, including:

[0229] According to the abnormal identification results of the power grid area where the IoT device is located, the power grid area with abnormalities is selected to construct the corresponding power grid abnormality positioning model, and the constructed power grid abnormality positioning model is optimized and solved to obtain the abnormal location information in the power grid area. The power grid abnormality positioning model simulates the power grid topology information in a digital way, and combines the abnormal identification results of the power grid area to perform abnormal positioning. The Markov decision process is the main implementation method of the power grid abnormality positioning model, in which the abnormal power grid area The abnormal location information determination process is as follows:

[0230] S31: Digitally monitor the grid area The grid topology information is simulated to obtain the grid area The topological structure diagram G=(E,V), where E represents the power grid area The location set of IoT devices in , V represents the power grid area The set of power grid lines between any two equipment locations in;

[0231] S32: Get the power grid area The power consumption spatiotemporal information of all IoT devices in the grid is collected, and the IoT devices are clustered based on the power consumption spatiotemporal information to obtain the power grid area The Q IoT device clusters in the network are calculated, the location mean of all IoT devices in each IoT device cluster is calculated, and the location mean is mapped in the power grid area. The grid line location with the closest mean value in the middle distance is taken as the grid area The Q initial abnormal position information in:

[0232] (Loc0(1),Loc0(2),...,Loc0(q),...,Loc0(Q))

[0233] in:

[0234] Loc0(q) represents the grid area The location mean mapping result of the qth IoT device cluster in the grid area The qth initial abnormal position information in ;

[0235] S33: Set the current iteration number of the initial abnormal location information to d, the initial value of d is 0, and the maximum value is Max, then the power grid area The d-th iteration result of the q-th initial abnormal location information in is Loc d (q), constructs the reward function of the initial abnormal position information iteration based on the Markov decision process, where the initial abnormal position information is obtained from Loc d (q) Iterate to Loc d The reward function for (q+1) is:

[0236] Reward(Loc d (q+1))=avg(Loc d (q+1))―avg(Loc d (q))

[0237] in:

[0238] Reward(Loc d (q+1)) indicates that the initial abnormal location information is from Loc d (q) Iterate to Loc d The reward function of (q+1);

[0239] avg(Loc d (q)) indicates the location Loc d (q) The average distance along the power grid line to the IoT device with abnormal spatiotemporal information of power consumption in the qth IoT device cluster;

[0240] S34: Iterate the Q initial abnormal position information based on the reward function, select the power line position that makes the reward function positive as the d+1th iteration result of the Q initial abnormal position information, until the power line position cannot be selected, and use the Q positions iterated at this time as the power grid area Abnormal location information.

[0241] S4: According to the abnormal position information, the inspection robot is driven to perform intelligent inspection based on virtual reality technology to obtain the power grid inspection results.

[0242] In step S4, the inspection robot is driven to perform intelligent inspection based on virtual reality technology according to the abnormal position information, including:

[0243] According to the abnormal position information, the inspection robot is driven to perform intelligent inspection based on virtual reality technology to obtain the power grid inspection results, where the power grid inspection results include the location of the damaged power grid line and the corresponding power grid line maintenance and processing steps. The intelligent inspection process is as follows:

[0244] S41: The inspection robot receives the position in the abnormal position information and arrives at the corresponding power grid line position;

[0245] S42: The inspection robot turns on a camera to obtain a video image of the power grid line location it has reached;

[0246] S43: The inspection robot transmits the acquired video image to the virtual reality device, and the virtual reality device performs three-dimensional virtual modeling on the video image to obtain a three-dimensional virtual scene of the power grid line location;

[0247] S44: comparing the three-dimensional virtual scene of the power grid line location with the three-dimensional scene of the normal power grid line. If there is a difference, it means that the power grid line location is damaged. The operator wears a virtual reality device to conduct further safety inspection of the three-dimensional virtual scene, and generates an operating device in the three-dimensional virtual scene.

[0248] S45: The operator uses the operating device to perform grid line maintenance processing on the grid line in the three-dimensional virtual scene in the virtual scene, and sends the grid line maintenance processing steps to the maintenance station closest to the grid line location. The maintenance station performs grid line maintenance according to the grid line maintenance processing steps.

[0249] Embodiment 2:

[0250] like Figure 2 , is a schematic diagram of the structure of an electronic device for implementing an intelligent inspection method based on the fusion of the Internet of Things and virtual reality based on an AI big model provided by an embodiment of the present invention.

[0251] The electronic device 1 may include a processor 10 , a memory 11 , a communication interface 13 and a bus, and may also include a computer program stored in the memory 11 and executable on the processor 10 , such as a program 12 .

[0252] The memory 11 includes at least one type of readable storage medium, and the readable storage medium includes a flash memory, a mobile hard disk, a multimedia card, a card-type memory (e.g., SD or DX memory, etc.), a magnetic memory, a disk, an optical disk, etc. The memory 11 may be an internal storage unit of the electronic device 1 in some embodiments, such as a mobile hard disk of the electronic device 1. The memory 11 may also be an external storage device of the electronic device 1 in other embodiments, such as a plug-in mobile hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device 1. Further, the memory 11 may also include both an internal storage unit of the electronic device 1 and an external storage device. The memory 11 may be used not only to store application software and various types of data installed in the electronic device 1, such as the code of the program 12, etc., but also to temporarily store data that has been output or is to be output.

[0253] The processor 10 may be composed of an integrated circuit in some embodiments, for example, a single packaged integrated circuit, or a plurality of integrated circuits with the same or different functions, including one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and combinations of various control chips. The processor 10 is the control core (Control Unit) of the electronic device, and uses various interfaces and lines to connect the various components of the entire electronic device, and executes or executes the program or module stored in the memory 11 (for realizing the program 12 of intelligent inspection based on the integration of the Internet of Things and virtual reality based on the AI ​​large model, etc.), and calls the data stored in the memory 11 to execute various functions of the electronic device 1 and process data.

[0254] The communication interface 13 may include a wired interface and / or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is generally used to establish a communication connection between the electronic device 1 and other electronic devices, and to achieve connection and communication between internal components of the electronic device.

[0255] The bus may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. The bus is configured to realize connection and communication between the memory 11 and at least one processor 10, etc.

[0256] Figure 2 Only an electronic device with components is shown, and those skilled in the art will understand that Figure 2 The structure shown does not constitute a limitation on the electronic device 1, and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.

[0257] For example, although not shown, the electronic device 1 may also include a power source (such as a battery) for supplying power to each component. Preferably, the power source may be logically connected to the at least one processor 10 through a power management device, so that the power management device can realize functions such as charging management, discharging management, and power consumption management. The power source may also include any components such as one or more DC or AC power sources, recharging devices, power failure detection circuits, power converters or inverters, power status indicators, etc. The electronic device 1 may also include a variety of sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be repeated here.

[0258] Optionally, the electronic device 1 may further include a user interface, which may be a display, an input unit (such as a keyboard), or a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, and an OLED (Organic Light-Emitting Diode) touch device. The display may also be appropriately referred to as a display screen or a display unit, which is used to display information processed in the electronic device 1 and to display a visual user interface.

[0259] It should be understood that the embodiment is for illustration only and the scope of the patent application is not limited to this structure.

[0260] The program 12 stored in the memory 11 in the electronic device 1 is a combination of multiple instructions. When running in the processor 10, it can achieve:

[0261] Using IoT sensors to collect IoT device serial data, transmitting the collected IoT device serial data to a backend server for preprocessing, and obtaining preprocessed IoT device serial data;

[0262] Construct a power grid anomaly recognition model to identify anomalies in the preprocessed IoT device sequence data;

[0263] According to the abnormal identification results of the power grid area where the IoT device is located, a power grid abnormality positioning model is constructed and optimized to obtain the abnormal location information in the power grid area;

[0264] According to the abnormal position information, the inspection robot is driven to perform intelligent inspection based on virtual reality technology to obtain the power grid inspection results.

[0265] Specifically, the specific implementation method of the processor 10 for the above instructions can refer to Figure 1 to Figure 2 The description of the relevant steps in the corresponding embodiments will not be repeated here.

[0266] It should be noted that the serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages and disadvantages of the embodiments. And the terms "including", "comprising" or any other variants thereof in this article are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "including a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0267] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0268] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. An intelligent inspection method based on the fusion of the Internet of Things and virtual reality based on an AI big model, characterized in that: The method comprises: S1: Using IoT sensors to collect IoT device sequence data, where the IoT device sequence data includes current, voltage and coordinate data of IoT devices, and transmitting the collected IoT device sequence data to a backend server for preprocessing to obtain preprocessed IoT device sequence data; S2: constructing a power grid anomaly recognition model to perform anomaly recognition on the preprocessed IoT device sequence data, wherein the power grid anomaly recognition model takes the preprocessed IoT device sequence data as input and takes the anomaly recognition result of the power grid area where the IoT device is located as output; S3: construct a power grid anomaly location model based on the anomaly identification results of the power grid area where the IoT device is located, and optimize and solve it to obtain the anomaly location information in the power grid area. The power grid anomaly location model simulates the power grid topology information in a digital way, and locates the anomaly in combination with the anomaly identification results of the power grid area; The grid anomaly location model simulates the grid topology information in a digital way and locates the anomaly in combination with the anomaly identification results of the grid area. The abnormal location information determination process is as follows: S31: Digitally identify areas of the power grid where anomalies exist The grid topology information is simulated to obtain the grid area The topological structure diagram G=(E,V), where E represents the abnormal power grid area The location set of IoT devices in , V represents the power grid area The set of power grid lines between any two equipment locations in; S32: Obtain the power grid area with abnormality The power consumption spatiotemporal information of all IoT devices in the network is collected, and the IoT devices are clustered based on the power consumption spatiotemporal information to obtain the power grid areas with abnormalities. The Q IoT device clusters in the cluster are calculated, the location mean of all IoT devices in each IoT device cluster is calculated, and the location mean is mapped to the abnormal power grid area. The grid line location with the closest mean value in the middle distance is regarded as the grid area with abnormality The Q initial abnormal position information in: (Loc0(1),Loc0(2),...,Loc0(q),...,Loc0(Q)) in: Loc0(q) indicates the abnormal power grid area The location mean mapping result of the qth IoT device cluster in , that is, the abnormal power grid area The qth initial abnormal position information in ; S33: Set the current iteration number of the initial abnormal location information to d, the initial value of d is 0, and the maximum value is Max, then there is an abnormal power grid area The d-th iteration result of the q-th initial abnormal location information in is Loc d (q), constructs the reward function of the initial abnormal position information iteration based on the Markov decision process, where the initial abnormal position information is obtained from Loc d (q) Iterate to Loc d The reward function for (q+1) is: Reward(Loc d (q+1))=avg(Loc d (q+1))-avg(Loc d (q)) in: Reward(Loc d (q+1)) indicates that the initial abnormal location information is from Loc d (q) Iterate to Loc d The reward function of (q+1); avg(Loc d (q)) indicates the location Loc d (q) The average distance along the power grid line to the IoT device with abnormal spatiotemporal information of power consumption in the qth IoT device cluster; S34: Iterate the Q initial abnormal position information based on the reward function, select the power line position that makes the reward function positive as the d+1th iteration result of the Q initial abnormal position information, until the power line position cannot be selected, and use the Q positions iterated at this time as the power grid area with abnormality Abnormal location information; S4: According to the abnormal position information, the inspection robot is driven to perform intelligent inspection based on virtual reality technology to obtain the power grid inspection results.

2. The intelligent inspection method based on the fusion of the Internet of Things and virtual reality based on the AI ​​big model as claimed in claim 1 is characterized in that: In the step S1, the IoT sensor is used to collect IoT device sequence data, including: The IoT sensor is used to collect IoT device serial data, where the IoT device serial data includes the current, voltage and coordinate data of the IoT device. The collected IoT device serial data is: {z n =(I n (t),U n (t),(x n ,y n ))|n∈[1,N]} I n (t)=(I n (t1),I n (t2),...,I n (t m ),...,I n (t M )) U n (t)=(U n (t1),U n (t2),...,U n (t m ),...,U n (t M )) in: z n Represents the current, voltage and coordinate data of the nth IoT device, I n (t) represents the current data of the nth IoT device, U n (t) represents the voltage data of the nth IoT device, t represents the timing information, (x n ,y n ) represents the plane coordinates of the nth IoT device, x n Indicates the plane coordinate value of the nth IoT device in the horizontal direction, y n Indicates the plane coordinate value of the nth IoT device in the vertical direction; I n (t m ) indicates that the nth IoT device senses the time t at the mth sensor m Current value, t M represents the current sensor sensing moment, and the time difference between adjacent sensor sensing moments is Δt; U n (t m ) indicates that the nth IoT device senses the time t at the mth sensor m Voltage value; The collected IoT device serial data is transmitted to the backend server for preprocessing to obtain the preprocessed IoT device serial data {Z n =(I′n(t),U′ n (t),(x n ,y n ))|n∈[1,N]},Z n Represents the preprocessed current, voltage and coordinate data of the nth IoT device, I′ n (t) represents the preprocessed current data of the nth IoT device, U′ n (t) represents the preprocessed voltage data of the nth IoT device.

3. The intelligent inspection method based on the fusion of the Internet of Things and virtual reality based on the AI ​​big model as claimed in claim 2 is characterized in that: The method of transmitting the collected IoT device serial data to the backend server for preprocessing includes: The collected IoT device serial data is transmitted to the backend server for preprocessing, where the current, voltage and coordinate data z of the nth IoT device in the IoT device serial data are n The preprocessing process is: S11: Opposite n Current data I n (t) Perform preprocessing to obtain the preprocessed current data I′ of the nth IoT device n (t): I′ n (t)=(I′ n (t1),I′ n (t2),...,I′ n (t m ),...,I′ n (t M )) in: I max Indicates the preset maximum current, I min Indicates the preset minimum current value; I′ n (t m ) represents the current value I n (t m )’s preprocessing results; S12: Opposite n Voltage data U n (t) Perform preprocessing to obtain the preprocessed voltage data I′ of the nth IoT device n (t): U′ n (t)=(U′ n (t1),U′ n (t2),...,U′ n (t m ),...,U′ n (t M )) in: U max Indicates the preset maximum voltage, U min Indicates the preset minimum voltage value; U′ n (t m ) represents the voltage value U n (t m )’s preprocessing results; S13: Constructs the pre-processed current, voltage and coordinate data of the nth IoT device: Z n =(I′ n (t),U′ n (t),(x n ,y n )) in: Z n Represents the preprocessed current, voltage, and coordinate data of the nth IoT device.

4. The intelligent inspection method based on the fusion of the Internet of Things and virtual reality based on the AI ​​big model as claimed in claim 1 is characterized in that: The step S2 constructs a power grid anomaly recognition model, including: Construct a power grid anomaly recognition model, and use the power grid anomaly recognition model to perform anomaly recognition on the preprocessed IoT device sequence data. The power grid anomaly recognition model takes the preprocessed IoT device sequence data as input and takes the anomaly recognition result of the power grid area where the IoT device is located as output. The power grid anomaly recognition model includes an input layer, a spatiotemporal feature mapping layer, a spatiotemporal feature correlation calculation layer, and an anomaly recognition layer; The input layer is used to divide the preprocessed IoT device sequence data according to the power grid area, and obtain the preprocessed current, voltage and coordinate data of all IoT devices in each power grid area; The spatiotemporal feature mapping layer is used to calculate the spatiotemporal features of each power grid area; The temporal and spatial feature correlation calculation layer is used to calculate the temporal and spatial feature correlation of each power grid area; The anomaly recognition layer is used to obtain anomaly recognition results for different power grid areas based on the spatiotemporal characteristics of the power grid area and the correlation of spatiotemporal characteristics.

5. The intelligent inspection method based on the fusion of the Internet of Things and virtual reality based on the AI ​​big model as claimed in claim 4 is characterized in that: The method of using the power grid anomaly recognition model to perform anomaly recognition on the preprocessed IoT device sequence data includes: The power grid anomaly recognition model is used to identify anomalies in the preprocessed IoT device sequence data, where the anomaly recognition process is as follows: S21: The input layer receives the preprocessed IoT device sequence data and divides the preprocessed IoT device sequence data according to the power grid area to obtain the preprocessed current, voltage and coordinate data of all IoT devices in each power grid area, where the power grid area Ω k The pre-processed current, voltage and coordinate data of all IoT devices are: in: Represents the power grid area Ω k Preprocessed current data of the sth IoT device Voltage data and coordinate data S k Represents the power grid area Ω k Total number of IoT devices in k∈[1,K], K represents the total number of power grid areas, Ω1,Ω2,...,Ω K represents K power grid areas; S22: The spatiotemporal feature mapping layer calculates the spatiotemporal features of each power grid area, where the power grid area Ω k The spatiotemporal characteristics are: in: W represents the feature mapping matrix, and T represents transpose; F k Represents the power grid area Ω k The spatiotemporal characteristics of Represents the power grid area Ω k The spatiotemporal information of power consumption of the s-th IoT device; exp(·) represents an exponential function with a natural constant as the base. Represents coordinate data Arrival at the power grid areaΩ k The shortest distance of the power grid equipment, including the step-up substation, step-down substation and distribution substation in the power grid area; Represents the power grid area Ω k The current timing characteristic value of the sth IoT device in, Represents the power grid area Ω k The voltage timing characteristic value of the s-th IoT device in; S23: The spatiotemporal feature correlation calculation layer calculates the spatiotemporal feature correlation of each power grid area, where the power grid area Ω k The spatial and temporal feature correlation of is: in: Per k Represents the power grid area Ω k The spatial and temporal characteristics of the correlation; Represents the power grid area Ω k The spatial and temporal feature correlation of the s-th IoT device in; ||·|| represents the L1 norm; S24: The anomaly recognition layer obtains anomaly recognition results of different power grid areas according to the spatiotemporal characteristics and the correlation of spatiotemporal characteristics of the power grid area, where the power grid area Ω k The abnormal recognition results are: in: ρ k Represents the power grid area Ω k The abnormal recognition result of k =1 indicates the power grid area Ω k There is an abnormality, ρ k =0 means the power grid area Ω k There are no anomalies; Represents the power grid area Ω k The spatiotemporal information of electricity consumption of the sth IoT device in If the L2 norm of k The sth IoT device is marked as an IoT device with abnormal spatiotemporal information of electricity consumption; ε represents the preset abnormal threshold.

6. The intelligent inspection method based on the fusion of the Internet of Things and virtual reality based on the AI ​​big model as claimed in claim 1 is characterized in that: In step S4, the inspection robot is driven to perform intelligent inspection based on virtual reality technology according to the abnormal position information, including: According to the abnormal position information, the inspection robot is driven to perform intelligent inspection based on virtual reality technology to obtain the power grid inspection results, where the power grid inspection results include the location of the damaged power grid line and the corresponding power grid line maintenance and processing steps. The intelligent inspection process is as follows: S41: The inspection robot receives the position in the abnormal position information and arrives at the corresponding power grid line position; S42: The inspection robot turns on a camera to obtain a video image of the power grid line location it has reached; S43: The inspection robot transmits the acquired video image to the virtual reality device, and the virtual reality device performs three-dimensional virtual modeling on the video image to obtain a three-dimensional virtual scene of the power grid line location; S44: comparing the three-dimensional virtual scene of the power grid line location with the three-dimensional scene of the normal power grid line. If there is a difference, it means that the power grid line location is damaged. The operator wears a virtual reality device to conduct further safety inspection of the three-dimensional virtual scene, and generates an operating device in the three-dimensional virtual scene. S45: The operator uses the operating device to perform grid line maintenance processing on the grid line in the three-dimensional virtual scene in the virtual scene, and sends the grid line maintenance processing steps to the maintenance station closest to the grid line location. The maintenance station performs grid line maintenance according to the grid line maintenance processing steps.

Images