A distributed intelligent terminal fault detection system and method based on Beidou positioning

Abstract

The application discloses a distributed intelligent terminal fault detection system and method based on Beidou positioning, and the system comprises a collection module, a fault detection module, a terminal positioning module and a navigation correction module.The collection module is used for collecting Beidou information of the distributed intelligent terminal.The fault detection module is used for collecting operation information of the distributed intelligent terminal, training a distributed intelligent terminal fault detection model based on a Transform architecture, and performing fault detection by using the distributed intelligent terminal fault detection model.The terminal positioning module is used for calculating coordinate information of the distributed intelligent terminal which has a fault according to a fault detection result, and performing positioning.The navigation correction module is used for correcting a distributed intelligent terminal Beidou navigation technology based on a multi-modal model according to a positioning result, and navigating to the distributed intelligent terminal which has a fault.The application effectively improves the fault detection level and operation and maintenance efficiency of the distributed intelligent terminal, and is beneficial to improving the safe operation level of a power distribution network.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of power equipment, and particularly relates to a distributed intelligent terminal fault detection system and method based on Beidou positioning. BACKGROUND

[0002] Effective detection of the faults of the distributed intelligent terminal and timely operation and maintenance can effectively improve the safe operation level of the power distribution network. However, the existing method still has the following deficiencies: 1) single data dependence, the traditional distributed intelligent terminal fault detection method usually depends on a single data source, which may lead to the limitation of fault detection, especially in the face of complex and variable environments, the characteristics of the fault cannot be fully captured, and the fault may be missed or misdetected; 2) poor real-time performance, the traditional fault detection method is often difficult to realize real-time monitoring, especially in the case of large data volume and complex analysis algorithm, the real-time performance of fault detection may be affected, so that the fault cannot be discovered and responded in time; 3) communication limitation, the distributed intelligent terminal usually transmits fault information to the operation and maintenance center through a wireless communication network (such as Wi-Fi, cellular network, etc.). However, in the case of poor network coverage or communication interruption, the fault information may not be transmitted in time, so that the operation and maintenance center cannot respond in time; 4) multi-path effect and signal interference, in the urban environment, due to the reflection and shielding of buildings, the distributed intelligent terminal may be affected by the multi-path effect, which increases the positioning error. In addition, the terminal may also be interfered by the signals of other wireless devices, further reducing the accuracy of fault detection and positioning. SUMMARY

[0003] In order to solve the above problems, the application provides a distributed intelligent terminal fault detection system and method based on Beidou positioning, which can improve the fault detection level of the distributed intelligent terminal and accurately locate the distributed intelligent terminal that has failed.

[0004] In order to achieve the above purpose, the application is realized by the following technical scheme:

[0005] The distributed intelligent terminal fault detection system based on Beidou positioning provided by the application comprises:

[0006] The acquisition module is used for constructing a pseudo-range model between the Beidou satellite and the distributed intelligent terminal, calculating the pseudo-range between a single Beidou satellite and the distributed intelligent terminal, locking the position information of the distributed intelligent terminal by using the pseudo-range model between multiple Beidou satellites and the distributed intelligent terminal, constructing a Beidou satellite geographic coordinate conversion model, converting the position information of the distributed intelligent terminal into geographic coordinates, and realizing the acquisition of the Beidou information of the distributed intelligent terminal;

[0007] The fault detection module is used to collect the operating information of the distributed intelligent terminal, train the distributed intelligent terminal fault detection model based on the Transformer architecture, and use the distributed intelligent terminal fault detection model to perform fault detection.

[0008] The terminal positioning module is used to calculate the coordinate information of the faulty distributed intelligent terminal based on the fault detection results and to locate it.

[0009] The navigation correction module is used to navigate to the malfunctioning distributed intelligent terminal based on the positioning results and the BeiDou navigation technology of the distributed intelligent terminal, which is corrected by a multimodal model.

[0010] A further improvement of the present invention is that the specific operations performed by the acquisition module include:

[0011] Construct a pseudorange model between BeiDou satellites and distributed intelligent terminals, expressed as:

[0012]

[0013] Where, ρ i Let X be the pseudorange of the i-th Beidou satellite and the distributed intelligent terminal. i ,Y i Z i Let (x0, y0, z0) be the known position of the i-th Beidou satellite in the geocentric-ground-fixed coordinate system, (x0, y0, z0) be the position of the distributed intelligent terminal, c be the speed of light, and δt be the clock offset of the distributed intelligent terminal. i It's a satellite clock deviation;

[0014] Using a pseudorange model between multiple BeiDou satellites and a distributed intelligent terminal, the position (X0, Y0, Z0) and clock deviation δt of the distributed intelligent terminal are solved.

[0015] A BeiDou satellite geographic coordinate transformation model is constructed to convert the location information of distributed intelligent terminals into geographic coordinates. The expression of the BeiDou satellite geographic coordinate transformation model is as follows:

[0016]

[0017] Where φ is the latitude of the distributed intelligent terminal, λ is the longitude of the distributed intelligent terminal, h is the altitude of the distributed intelligent terminal, and R E is the average radius of the Earth.

[0018] A further improvement of the present invention is that the specific operations performed by the fault detection module include:

[0019] Collect operational information from distributed intelligent terminals, specifically including:

[0020] The collected operation information of the distributed intelligent terminal includes temperature, voltage, current and power, and the expression is:

[0021] X(t) = [T(t), V(t), I(t), P(t)] (3);

[0022] Wherein, X(t) is the operation information of the distributed intelligent terminal at time t, T(t) is the temperature collected at time t, V(t) is the voltage collected at time t, I(t) is the current collected at time t, and P(t) is the power collected at time t;

[0023] The input sample composed of the operation information of the distributed intelligent terminal is used to train the distributed intelligent terminal fault detection model, wherein the distributed intelligent terminal fault detection model includes a position encoding model of a Transformer architecture, a self-attention model, a multi-head attention feature extraction model and a distributed intelligent terminal fault prediction model, and specifically includes:

[0024] The expression of the input sample composed of the operation information of the distributed intelligent terminal is:

[0025]

[0026] Wherein, X is the input sample of the Transformer architecture, and X(N) is the operation information of the distributed intelligent terminal at time N;

[0027] The position encoding model of the Transformer architecture is constructed, and the expression of the position encoding model of the Transformer architecture is:

[0028]

[0029]

[0030] X input = X + PE (7);

[0031] Wherein, PE is position encoding information, t is time step, g is dimension index, d is total dimension of features, and X input is an input matrix with position encoding;

[0032] The self-attention model is constructed to calculate the mutual relationship between different time steps, and the expression is:

[0033]

[0034]

[0035] Wherein: Q, K and V are query, key and value vector models respectively, W Q , W K, W V are the weights of query, key and value vector model respectively, Attention(Q,K,V) is attention weight, d k is the dimension of key vector, K T is the relationship coefficient between core key and related target in self-attention mechanism.

[0036] The multi-head attention feature extraction model is constructed, and the expression is:

[0037]

[0038] Wherein: MultiHead(Q,K,V) is a multi-head attention feature extraction model, Concat() is a string concatenation function, head j is the attention weight of the jth sample, Q j , K j , V j are the query, key and value vector model of the jth sample respectively.

[0039] The distributed intelligent terminal fault prediction model based on the Transformer architecture is constructed, and the expression is:

[0040] y=σ(W out h+b out ) (11);

[0041] Wherein, y is the distributed intelligent terminal fault prediction result, h is the feature vector processed by the Transformer architecture, W out and b out are the weights and bias of the output layer, and sigma is an activation function.

[0042] According to the output distributed intelligent terminal fault prediction result y, a threshold value tau is set to judge whether there is a fault, and an early warning is issued, if the distributed intelligent terminal fault prediction result y is greater than the threshold value tau, then output 1, judge that the distributed intelligent terminal exists fault, the expression is:

[0043]

[0044] The further improvement of the application is that the distributed intelligent terminal Beidou navigation technology based on multi-modal model correction comprises:

[0045] Based on the basic coordinate system and the state vector, the relative position and the relative distributed intelligent terminal speed of the distributed intelligent terminal away from the fault are obtained, and the expression is:

[0046]

[0047] Wherein: Q z(t) is the state vector of the distributed intelligent terminal at time t that is relatively faulty.

[0048] r z (t)=[x z (t),y z (t),Z z (t)] T Let be the relative position of the distributed intelligent terminal at time t.

[0049] V z (t)=[v x (t),v y (t),v z (t)] T Let t be the speed relative to the distributed intelligent terminal at time t;

[0050] M z (t)=[M G (t),M H (t),M D [(t)] represents the multimodal data measured at time t, where M G (t) represents the satellite signal, indicating whether the distributed terminal is faulty; a value of 1 indicates a fault, and a value of 0 indicates no fault; M H (t) represents the space electromagnetic environment data, which affects signal transmission quality; M D (t) represents terrain information, including contour lines and terrain slope;

[0051] Construct a dynamic model, the expression of which is:

[0052] Q z (t+1)=FQ z (t)+Bu(t)+w(t) (14);

[0053] Among them: Q z (t+1) is the state vector of the distributed intelligent terminal relative to the fault at time t+1, F is the state transition matrix, B is the control matrix, u(t) is the control input, and w(t) is the process noise;

[0054] Construct a multimodal data fusion model, expressed as:

[0055]

[0056] Where: G(t+1) is the state updated at time t+1, r GNSS (t+1) represents the GNSS position, z vision (t+1) represents the visual features at time t+1, and H is the observation matrix;

[0057] A relative position and speed updating model fusing multi-modal data is constructed, and a relative position and speed of the distributed intelligent terminal is estimated, wherein an expression of the relative position and speed updating model fusing multi-modal data is:

[0058]

[0059] Wherein: is a state vector predicted at t+1,

[0060] is a relative position and speed of the distributed intelligent terminal at t+1 respectively, K(t+1) is a Kalman gain at t+1, H T is a transpose of an observation matrix, P(t+1|t) is a covariance matrix of a state estimation at t+1 at t, P(t|t-1) is a covariance matrix of a state estimation at t at t-1, R is an observation noise covariance, I is a unit matrix, values on the diagonal are 1, and values at other positions are 0.

[0061] The distributed intelligent terminal fault detection method based on Beidou positioning provided by the application comprises the following operations:

[0062] The Beidou information of the distributed intelligent terminal is collected, including constructing a pseudo-range model between the Beidou satellite and the distributed intelligent terminal, calculating the pseudo-range between a single Beidou satellite and the distributed intelligent terminal, locking the position information of the distributed intelligent terminal by using the pseudo-range model between multiple Beidou satellites and the distributed intelligent terminal, and constructing a Beidou satellite geographic coordinate conversion model to convert the position information of the distributed intelligent terminal into geographic coordinates.

[0063] The running information of the distributed intelligent terminal is collected, and a distributed intelligent terminal fault detection model is trained based on a Transformer architecture, and the distributed intelligent terminal fault detection model is used for fault detection.

[0064] If the fault detection result is a fault, a fault signal is sent to the Beidou navigation satellite, the Beidou navigation satellite receives the signal, and the coordinate information of the distributed intelligent terminal with the fault is calculated, the coordinate information of the distributed intelligent terminal with the fault is analyzed into longitude, latitude and height coordinates by the operation and maintenance center.

[0065] The Beidou navigation technology of the distributed intelligent terminal based on the multi-modal model correction is used to navigate to the distributed intelligent terminal with the fault.

[0066] The further improvement of the application is that the Beidou information of the distributed intelligent terminal is collected, specifically including:

[0067] The pseudo-range model between the Beidou satellite and the distributed intelligent terminal is constructed, and the expression is:

[0068]

[0069] wherein, ρ i is the measured pseudo-range between the i-th Beidou satellite and the distributed intelligent terminal, (X i , Y i , Z i ) is the known position of the i-th Beidou satellite in the Earth-Centered Earth-Fixed coordinate system, (X0, Y0, Z0) is the position of the distributed intelligent terminal, c is the speed of light, δt is the clock bias of the distributed intelligent terminal, δt i is the satellite clock bias;

[0070] The position (X0, Y0, Z0) of the distributed intelligent terminal and the clock bias δt of the distributed intelligent terminal are solved by using the pseudo-range model between multiple Beidou satellites and the distributed intelligent terminal.

[0071] A Beidou satellite geographic coordinate conversion model is constructed to convert the position information of the distributed intelligent terminal into geographic coordinates, wherein the expression of the Beidou satellite geographic coordinate conversion model is:

[0072]

[0073] wherein, φ is the latitude of the distributed intelligent terminal, λ is the longitude of the distributed intelligent terminal, h is the altitude of the distributed intelligent terminal, R E is the average radius of the earth.

[0074] The further improvement of the present application is that the running information of the distributed intelligent terminal is collected, and a distributed intelligent terminal fault detection model is trained based on the Transformer architecture, and the distributed intelligent terminal fault detection model is used for fault detection, specifically including:

[0075] The running information of the distributed intelligent terminal is collected, wherein the collected running information of the distributed intelligent terminal includes temperature, voltage, current and power, and the expression is:

[0076] X(t) = [T(t), V(t), I(t), P(t)] (3);

[0077] wherein, X(t) is the running information of the distributed intelligent terminal at time t, T(t) is the temperature collected at time t, V(t) is the voltage collected at time t, I(t) is the current collected at time t, and P(t) is the power collected at time t.

[0078] The input sample composed of the running information of the distributed intelligent terminal is used to train the distributed intelligent terminal fault detection model, wherein the distributed intelligent terminal fault detection model comprises a position encoding model of a Transformer architecture, a self-attention model, a multi-head attention feature extraction model and a distributed intelligent terminal fault prediction model, and specifically comprises:

[0079] The expression of the input sample composed of the running information of the distributed intelligent terminal is:

[0080]

[0081] Wherein, X is the input sample of the Transformer architecture, and X(N) is the running information of the distributed intelligent terminal at N time;

[0082] The position encoding model of the Transformer architecture is constructed, and the expression of the position encoding model of the Transformer architecture is:

[0083]

[0084]

[0085] X input =X+PE (7);

[0086] Wherein, PE is position encoding information, t is a time step, g is a dimension index, d is the total dimension of the feature, and X input is an input matrix with position encoding added;

[0087] The self-attention model is constructed to calculate the mutual relationship between different time steps, and the expression is:

[0088]

[0089]

[0090] Wherein: Q, K and V are query, key and value vector models respectively, W Q , W K and W V are weights of the query, key and value vector models respectively, Attention(Q,K,V) is an attention weight, d k is the dimension of the key vector, and K T is the relationship coefficient between the core key and the related target in the self-attention mechanism;

[0091] The multi-head attention feature extraction model is constructed, and the expression is:

[0092]

[0093] Wherein: MultiHead (Q, K, V) is a multi-head attention feature extraction model, Concat () is a string concatenation function, head j is the attention weight of the jth sample, Q j , K j , V j Respectively, the query, key and value vector model of the jth sample;

[0094] A distributed intelligent terminal fault prediction model based on the Transformer architecture is constructed, and the expression is:

[0095] y=σ (W out h+b out ) (11);

[0096] Wherein, y is the distributed intelligent terminal fault prediction result, h is the feature vector processed by the Transformer architecture, W out and b out are the weights and bias of the output layer, and sigma is an activation function;

[0097] According to the output of the distributed intelligent terminal fault prediction result y, set the threshold value tau to judge whether there is a fault, and issue a warning, if the distributed intelligent terminal fault prediction result y is greater than the threshold value tau, output 1, judge that the distributed intelligent terminal exists fault, the expression is:

[0098]

[0099] The further improvement of the application is that the distributed intelligent terminal Beidou navigation technology based on multi-modal model correction comprises:

[0100] Based on the basic coordinate system and the state vector, the relative position and the relative distributed intelligent terminal speed of the distributed intelligent terminal away from the fault are obtained, and the expression is:

[0101]

[0102] Wherein: Q z (t) is the state vector of the relative distributed intelligent terminal away from the fault at t,

[0103] r z (t)=[x z (t),y z (t),Z z (t)] T Is the relative position away from the distributed intelligent terminal at t,

[0104] V z (t)=[v x (t),vy (t),v z (t)] T is the velocity of the distributed intelligent terminal at time t;

[0105] M z (t)=[M G (t),M H (t),M D (t)] is the multi-modal data measured at time t, wherein M G (t) is the satellite signal, indicating whether the distributed terminal has a fault, and the value of 1 indicates that there is a fault, and the value of 0 indicates that there is no fault; M H (t) is the spatial electromagnetic environment data, which affects the signal transmission quality; M D (t) is the terrain information, including contour lines, terrain slope, etc.

[0106] A dynamic model is constructed, and the expression is:

[0107] Q z (t+1)=FQ z (t)+Bu(t)+w(t) (14);

[0108] wherein: Q z (t+1) is the state vector of the distributed intelligent terminal relative to the fault at time t+1, F is the state transition matrix, B is the control matrix, u(t) is the control input, and w(t) is the process noise;

[0109] A multi-modal data fusion model is constructed, and the expression is:

[0110]

[0111] wherein: G(t+1) is the updated state at time t+1, r GNSS (t+1) is the GNSS position, z vision (t+1) is the visual feature at time t+1, and H is the observation matrix;

[0112] A relative position and velocity update model of fused multi-modal data is constructed, and the relative position and velocity relative to the distributed intelligent terminal are estimated, wherein the expression of the relative position and velocity update model of fused multi-modal data is:

[0113]

[0114] wherein: is the predicted state vector at time t+1,

[0115] Let K(t+1) be the relative position and relative velocity of the distance from the distributed intelligent terminal at time t+1, and H be the Kalman gain at time t+1. T Let P(t+1|t) be the covariance matrix of the state estimate at time t+1, P(t|t-1) be the covariance matrix of the state estimate at time t-1, R be the observation noise covariance, and I be the identity matrix, with values ​​of 1 on the diagonal and 0 elsewhere.

[0116] The beneficial effects of this invention are: by using a distributed intelligent terminal fault detection model based on the Transformer architecture, this invention improves the overall scope and accuracy of distributed intelligent terminal fault detection.

[0117] This invention proposes a distributed terminal BeiDou navigation technology based on multimodal model correction, which realizes dynamic correction of the relative position and velocity of distributed intelligent terminals, thereby improving the positioning accuracy and navigation performance of distributed intelligent terminals. Attached Figure Description

[0118] Figure 1 This is a schematic diagram of the system principle in an embodiment of the present invention;

[0119] Figure 2 This is a flowchart of the method in an embodiment of the present invention. Detailed Implementation

[0120] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0121] like Figure 2 As shown in this embodiment, a distributed intelligent terminal fault detection method based on BeiDou positioning includes the following steps:

[0122] Step 1, collect the Beidou information of the distributed intelligent terminal. The reference coordinate system used by the Beidou satellite navigation system is usually the geocentric and fixed coordinate system. In this embodiment, the longitude, latitude and altitude of the distributed intelligent terminal are parsed from several Beidou satellites by constructing a model. Specifically, it includes: constructing a pseudo-range model between Beidou satellites and the distributed intelligent terminal, calculating the pseudo-range between a single Beidou satellite and the distributed intelligent terminal; using the pseudo-range model between multiple Beidou satellites and the distributed intelligent terminal to lock the position information of the distributed intelligent terminal; constructing a Beidou satellite geographic coordinate conversion model to convert the position information of the distributed intelligent terminal into geographic coordinates, including longitude, latitude and altitude, and realizing the collection of the Beidou information of the distributed intelligent terminal.

[0123] Step 2, collect the running information of the distributed intelligent terminal, including temperature, voltage, current, power and other information, train the distributed intelligent terminal fault detection model based on the Transformer architecture, the distributed intelligent terminal fault detection model includes the position encoding model, the self-attention model, the multi-head attention feature extraction model and the Transformer-based distributed intelligent terminal fault prediction model of the Transformer architecture, according to the predicted fault result, set the threshold value, and judge whether the distributed intelligent terminal has a fault.

[0124] Step 3, if the fault detection result is a fault, send a fault signal to the Beidou navigation satellite, the Beidou navigation satellite receives the signal, measures the pseudo-range according to expressions (1)-(3) and calculates the coordinate information of the distributed intelligent terminal with a fault, the operation and maintenance center parses the coordinate information of the distributed intelligent terminal with a fault into longitude, latitude and altitude coordinates, and realizes the accurate positioning of the fault distributed intelligent terminal.

[0125] Step 4, based on the multi-modal model corrected distributed intelligent terminal Beidou navigation technology, navigate to the distributed intelligent terminal with a fault. In this embodiment, a plurality of data sources (such as satellite signal data, distributed intelligent terminal position data, space electromagnetic environment data, terrain information, etc.) are trained and analyzed by using a multi-modal model, based on the relative position and speed of the distributed intelligent terminal at the previous moment, the relative position and speed of the distributed intelligent terminal at the next moment are predicted, the dynamic correction of the relative position and speed of the distributed intelligent terminal is realized, and the positioning accuracy and navigation performance of the distributed intelligent terminal are improved.

[0126] In step 1, the reference coordinate system used by the Beidou satellite navigation system is usually the geocentric and fixed coordinate system. A pseudo-range model between Beidou satellites and the distributed intelligent terminal is constructed, and the pseudo-range model represents the clock error between Beidou satellites and the distributed intelligent terminal, and the expression is:

[0127]

[0128] wherein ρ i is the measured pseudo-range between the i-th Beidou satellite and the distributed intelligent terminal, (X i , Y i , Z i ) is the known position of the i-th Beidou satellite in the Earth-Centered Earth-Fixed coordinate system, (X0, Y0, Z0) is the position of the distributed intelligent terminal, c is the speed of light, δt is the clock bias of the distributed intelligent terminal, and δt i is the clock bias of the satellite.

[0129] To solve the position (X0, Y0, Z0) of the distributed intelligent terminal, the pseudo-range model between multiple Beidou satellites and the distributed intelligent terminal is used for solving. The present embodiment uses four satellites (i = 1, 2, 3, 4) to establish the non-linear equation group shown in expression (2) to solve the position of the distributed intelligent terminal.

[0130]

[0131] wherein ρ1, ρ2, ρ3, ρ4 are the measured pseudo-ranges between the 1st, 2nd, 3rd, and 4th Beidou satellites and the distributed intelligent terminal, respectively, (X1, Y1, Z1), (X2, Y2, Z2), (X3, Y3, Z3), (X4, Y4, Z4) are the known positions of the 1st, 2nd, 3rd, and 4th Beidou satellites in the Earth-Centered Earth-Fixed coordinate system, respectively, and δt1, δt2, δt3, δt4 are the clock biases of the 1st, 2nd, 3rd, and 4th Beidou satellites, respectively.

[0132] By solving the non-linear equation group of expression (2), the three-dimensional coordinates (X0, Y0, Z0) and the clock bias δt of the distributed intelligent terminal are obtained. The present embodiment constructs a Beidou satellite geographic coordinate conversion model, as shown in expression (3), and converts the position information of the distributed intelligent terminal into geographic coordinates according to expression (3) to obtain the longitude, latitude, and altitude of the distributed terminal.

[0133]

[0134] wherein φ is the latitude of the distributed intelligent terminal, λ is the longitude of the distributed intelligent terminal, h is the altitude of the distributed intelligent terminal, and R E is the average radius of the Earth.

[0135] In step 2, the collected operation information of the distributed intelligent terminal includes temperature, voltage, current, and power, and the expression is:

[0136] X(t) = [T(t), V(t), I(t), P(t)] (4);

[0137] Wherein, X(t) is the running information of the distributed intelligent terminal at t time, T(t) is the temperature collected at t time, V(t) is the voltage collected at t time, I(t) is the current collected at t time, and P(t) is the power collected at t time.

[0138] The expression of the input sample composed of the running information of the distributed intelligent terminal is:

[0139]

[0140] Wherein, X is the input sample of the Transformer architecture, X(N) is the running information of the distributed intelligent terminal at N time, and N is the length of the time sequence.

[0141] The position encoding model of the Transformer architecture is constructed. Since the Transformer architecture does not have intrinsic time sequence information, the position information of the time sequence needs to be embedded into the input data through position encoding. The expression of the position encoding model of the Transformer architecture is:

[0142]

[0143]

[0144] X input =X+PE (8);

[0145] Wherein, PE is the position encoding information, t is the time step, g is the dimension index, d is the total dimension of the feature, and X input is the input matrix with position encoding.

[0146] The self-attention model is constructed to calculate the mutual relationship between different time steps, and the expression is:

[0147]

[0148]

[0149] Wherein: Q, K, and V are query, key and value vector models respectively, W Q , W K , and W V are the weights of the query, key and value vector models respectively, Attention(Q, K, V) is the attention weight, d k is the dimension of the key vector, and K T is the relationship coefficient between the core key and the related target in the self-attention mechanism.

[0150] The multi-head attention feature extraction model is constructed, and the multi-head attention feature extraction model of the Transformer architecture can capture the patterns of different features in the data. In this embodiment, the features of the distributed intelligent terminal are extracted by the multi-head attention feature extraction model. The expression is:

[0151]

[0152] Wherein: MultiHead(Q, K, V) is a multi-head attention feature extraction model, Concat() is a string concatenation function, head j is the attention weight of the jth sample, Q j , K j , and V j are the query, key, and value vector models of the jth sample, respectively.

[0153] The distributed intelligent terminal fault prediction model based on the Transformer architecture is constructed, and the expression is:

[0154] y = σ(W out h + b out ) (12);

[0155] Wherein, y is the distributed intelligent terminal fault prediction result, h is the feature vector processed by the Transformer architecture, W out and b out are the weights and biases of the output layer, and σ is the activation function.

[0156] According to the output distributed intelligent terminal fault prediction result y, a threshold τ is set to determine whether there is a fault, and a warning is issued. If the distributed intelligent terminal fault prediction result y is greater than the threshold τ, 1 is output, and it is judged that the distributed intelligent terminal has a fault. The expression is:

[0157]

[0158] In step 4, the proposed distributed terminal Beidou navigation technology based on multi-modal model correction is shown in expressions (14) to (17).

[0159] Expression (14) is the basic coordinate system and state vector, including the relative position of the distributed intelligent terminal to the fault and the relative speed of the distributed intelligent terminal.

[0160]

[0161] Wherein: Q z (t) is the state vector of the distributed intelligent terminal relative to the fault at time t,

[0162] r z (t) = [xz (t), y z (t), Z z (t)] T is the relative position of the distributed intelligent terminal at time t,

[0163] V z (t) = [v x (t), v y (t), v z (t)] T is the velocity of the distributed intelligent terminal relative to the distributed intelligent terminal at time t;

[0164] M z (t) = [M G (t), M H (t), M D (t)] is the measured multi-modal data at time t, where M G (t) is the satellite signal, indicating whether the distributed terminal has a fault, taking a value of 1 indicating a fault, and a value of 0 indicating no fault; M H (t) is spatial electromagnetic environment data, affecting signal transmission quality; M D (t) is terrain information, including contour lines, terrain slope, etc.

[0165] A dynamic model is constructed, and inertial measurement unit (IMU) data is used to describe the dynamic equation of motion, with the expression being:

[0166] Q z (t+1) = FQ z (t) + Bu(t) + w(t) (15);

[0167] where: Q z (t+1) is the state vector of the distributed intelligent terminal relative to the fault at time t+1, F is the state transition matrix, B is the control matrix, u(t) is the control input (such as acceleration and IMU data), w(t) is the process noise, representing the uncertainty of the model.

[0168] A multi-modal data fusion model is constructed, and data from multiple modal sensors (such as vision, GNSS, etc.) is fused to update the state estimation. The expression is:

[0169]

[0170] where: G(t+1) is the updated state at time t+1, r GNSS (t+1) is the GNSS position, z vision (t+1) is the visual feature at time t+1, and H is the observation matrix.

[0171] A relative position and speed updating model fusing multi-modal data is constructed, and an expression of the relative position and speed updating model fusing multi-modal data is as follows:

[0172]

[0173] wherein: is a state vector predicted at t+1, is a relative position and a relative speed of the distance distributed intelligent terminal at t+1 respectively, K(t+1) is a Kalman gain at t+1, H T is a transpose of an observation matrix, P(t+1|t) is a covariance matrix of a state estimation at t+1 at t, P(t|t-1) is a covariance matrix of a state estimation at t at t-1, R is an observation noise covariance, I is a unit matrix, values on the diagonal are 1, and values at other positions are 0.

[0174] The relative position and the relative speed of the distance distributed intelligent terminal are estimated through the above steps Autonomous navigation of the fault distributed intelligent terminal is realized.

[0175] The above method is realized through a distributed intelligent terminal fault detection system based on Beidou positioning of the embodiment, as shown in Figure 1 The system comprises:

[0176] The acquisition module is configured to construct a pseudo-range model between the Beidou satellite and the distributed intelligent terminal, calculate the pseudo-range between a single Beidou satellite and the distributed intelligent terminal, lock the position information of the distributed intelligent terminal by using the pseudo-range model between multiple Beidou satellites and the distributed intelligent terminal, construct a Beidou satellite geographic coordinate conversion model, convert the position information of the distributed intelligent terminal into geographic coordinates, and realize acquisition of the Beidou information of the distributed intelligent terminal.

[0177] The fault detection module is configured to acquire the running information of the distributed intelligent terminal, train a distributed intelligent terminal fault detection model based on a Transformer architecture, and perform fault detection by using the distributed intelligent terminal fault detection model.

[0178] The terminal positioning module is configured to calculate the coordinate information of the distributed intelligent terminal that has occurred a fault according to the fault detection result, and perform positioning.

[0179] The navigation correction module is configured to correct the Beidou navigation technology of the distributed intelligent terminal based on a multi-modal model according to the positioning result, and navigate to the distributed intelligent terminal that has occurred a fault.

[0180] As used herein, and unless otherwise indicated, all technical and scientific terms have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. As used herein, except where the context demands otherwise, the use of the singular includes the plural. As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items. As used herein, the term "or" means and / or. As used herein, the term "comprises the steps of" includes the steps of any combination of the recited steps.

[0181] The above description is further explained with reference to the specific embodiments. The specific embodiments are provided to explain the present application and to demonstrate at least one of the best modes contemplated for carrying out the application. Of course, the application is not necessarily limited to the particular embodiments described. It is apparent that various modifications and changes can be made to the present application without departing from the spirit and scope of the application.

Claims

1. A distributed intelligent terminal fault detection system based on Beidou positioning, characterized in that: The system comprises: The acquisition module is configured to construct a pseudo-range model between the Beidou satellite and the distributed intelligent terminal, calculate the pseudo-range between a single Beidou satellite and the distributed intelligent terminal, lock the position information of the distributed intelligent terminal by using the pseudo-range model between multiple Beidou satellites and the distributed intelligent terminal, construct a Beidou satellite geographic coordinate conversion model, and convert the position information of the distributed intelligent terminal into geographic coordinates to realize the acquisition of the Beidou information of the distributed intelligent terminal; The fault detection module is configured to acquire the operation information of the distributed intelligent terminal, train a distributed intelligent terminal fault detection model based on a Transformer architecture, and perform fault detection by using the distributed intelligent terminal fault detection model; The terminal positioning module is configured to calculate the coordinate information of the distributed intelligent terminal that has failed according to the fault detection result, and perform positioning. The navigation correction module is configured to correct the Beidou navigation technology of the distributed intelligent terminal based on a multi-modal model according to the positioning result, and navigate to the distributed intelligent terminal that has failed. The Beidou navigation technology of the distributed intelligent terminal corrected based on the multi-modal model specifically comprises: Based on the basic coordinate system and the state vector, the relative position and the relative speed of the distributed intelligent terminal that has failed are obtained, and the expression is as follows: where: Q z (t) is the state vector of the distributed intelligent terminal relative to the fault at time t, r z (t) = [x z (t),y z (t),Z z (t)] T is the relative position of the distributed intelligent terminal at time t, V z (t) = [v x (t),v y (t),v z (t)] T velocity of the distributed intelligent terminal at time t M z (t) = [M G (t), M H (t), M D (t)] is the multi-modal data measured at time t, where M G (t) is the satellite signal, indicating whether the distributed terminal has a fault, taking the value of 1 to indicate that there is a fault, and taking the value of 0 to indicate that there is no fault; M H (t) is the spatial electromagnetic environment data, which affects the signal transmission quality; M D (t) is the terrain information, including contour lines and terrain slope; A dynamic model is constructed, and the expression is as follows: Q z (t+1) = FQ z (t) + Bu(t) + w(t) (14); wherein: Q z (t+1) is the state vector of the distributed intelligent terminal relative to the fault at time t+1, F is the state transition matrix, B is the control matrix, u(t) is the control input, and w(t) is the process noise. A multi-modal data fusion model is constructed, and the expression is as follows: wherein: G(t+1) is the updated state at time t+1, r GNSS (t+1) is the GNSS position, z vision (t+1) is the visual feature at time t+1, and H is the observation matrix. A relative position and speed updating model of the fused multi-modal data is constructed, and the relative position and speed of the distributed intelligent terminal are estimated, wherein the expression of the relative position and speed updating model of the fused multi-modal data is as follows: wherein: is a state vector predicted at time t+1, is a relative position and a relative velocity of the intelligent terminal from the intelligent terminal at time t+1, respectively, K(t+1) is a Kalman gain at time t+1, H T is a transpose of an observation matrix, P(t+1|t) is a covariance matrix of a state estimation at time t+1 at time t, P(t|t-1) is a covariance matrix of a state estimation at time t at time t-1, R is an observation noise covariance, and I is an identity matrix having a value of 1 on a diagonal and a value of 0 at other positions.

2. The distributed intelligent terminal fault detection system based on Beidou positioning according to claim 1, characterized in that: The acquisition module performs the following specific operations: A pseudo-range model between the Beidou satellite and the distributed intelligent terminal is constructed, and the expression is as follows: Where, ρ i Let X be the pseudorange of the i-th Beidou satellite and the distributed intelligent terminal. i ,Y i Z i Let (x0, y0, z0) be the known position of the i-th Beidou satellite in the geocentric-ground-fixed coordinate system, (x0, y0, z0) be the position of the distributed intelligent terminal, c be the speed of light, and δt be the clock offset of the distributed intelligent terminal. i It's a satellite clock deviation; The position (X0, Y0, Z0) of the distributed intelligent terminal and the clock deviation δt of the distributed intelligent terminal are solved by using the pseudo-range model between multiple Beidou satellites and the distributed intelligent terminal; A Beidou satellite geographic coordinate conversion model is constructed to convert the position information of the distributed intelligent terminal into geographic coordinates, and the expression of the Beidou satellite geographic coordinate conversion model is as follows: wherein φ is the latitude of the distributed intelligent terminal, λ is the longitude of the distributed intelligent terminal, h is the altitude of the distributed intelligent terminal, R E is the average radius of the Earth. 3.The distributed intelligent terminal fault detection system based on Beidou positioning of claim 1, wherein: The fault detection module performs the following specific operations: The operation information of the distributed intelligent terminal is acquired, and specifically includes: The acquired operation information of the distributed intelligent terminal includes temperature, voltage, current, and power, and the expression is as follows: X(t) = [T(t), V(t), I(t), P(t)] (3); wherein X(t) is the operation information of the distributed intelligent terminal at time t, T(t) is the temperature acquired at time t, V(t) is the voltage acquired at time t, I(t) is the current acquired at time t, and P(t) is the power acquired at time t; The distributed intelligent terminal fault detection model is trained by using input samples composed of the operation information of the distributed intelligent terminal, wherein the distributed intelligent terminal fault detection model includes a position encoding model, a self-attention model, a multi-head attention feature extraction model, and a distributed intelligent terminal fault prediction model of the Transformer architecture, and specifically includes: The expression of the input samples composed of the operation information of the distributed intelligent terminal is as follows: Wherein, X is the input sample of the Transformer architecture, and X(N) is the running information of the distributed intelligent terminal at N time; A position encoding model of the Transformer architecture is constructed, and an expression of the position encoding model of the Transformer architecture is: X input = X + PE (7); where PE is the position encoding information, t is the time step, g is the dimension index, d is the total dimension of the feature, X input is the input matrix with position encodings added; A self-attention model is constructed to calculate the mutual relationship between different time steps, and an expression is: wherein: Q, K, V are query, key and value vector models respectively, W Q , W K , W V are weights of the query, key and value vector models respectively, Attention(Q, K, V) is an attention weight, d k is a dimension of the key vector, K T is a relationship coefficient between the core key and the relevant target in the self-attention mechanism, and softmax() is a normalization exponential function. A multi-head attention feature extraction model is constructed, and an expression is: wherein: MultiHead(Q, K, V) is a multi-head attention feature extraction model, Concat() is a string concatenation function, head j is the attention weight of the jth sample, Q j , K j , V j are the query, key and value vector models of the jth sample, respectively; A distributed intelligent terminal fault prediction model based on the Transformer architecture is constructed, and an expression is: y = σ(W out h + b out ) (11) wherein y is the distributed intelligent terminal fault prediction result, h is the feature vector processed through the Transformer architecture, W out and b out are the weights and bias of the output layer, and σ is the activation function. According to the output distributed intelligent terminal fault prediction result y, a threshold value tau is set to determine whether a fault exists, and a warning is issued, if the distributed intelligent terminal fault prediction result y is greater than the threshold value tau, then output 1, to determine that the distributed intelligent terminal has a fault, and an expression is:

4. A distributed intelligent terminal fault detection method based on Beidou positioning, characterized in that: The method comprises the following operations: Collecting Beidou information of the distributed intelligent terminal, including constructing a pseudo-range model between the Beidou satellite and the distributed intelligent terminal, calculating the pseudo-range between a single Beidou satellite and the distributed intelligent terminal, locking the position information of the distributed intelligent terminal by using the pseudo-range model between multiple Beidou satellites and the distributed intelligent terminal, and constructing a Beidou satellite geographic coordinate conversion model to convert the position information of the distributed intelligent terminal into geographic coordinates; Collecting the running information of the distributed intelligent terminal, and training a distributed intelligent terminal fault detection model based on the Transformer architecture, and performing fault detection by using the distributed intelligent terminal fault detection model; If the fault detection result is a fault, a fault signal is sent to the Beidou navigation satellite, the Beidou navigation satellite receives the signal, and calculates the coordinate information of the distributed intelligent terminal that has a fault, and the operation and maintenance center analyzes the coordinate information of the distributed intelligent terminal that has a fault and converts it into longitude, latitude and height coordinates; Based on the multi-modal model correction distributed intelligent terminal Beidou navigation technology, the navigation is to the distributed intelligent terminal that has a fault; The multi-modal model correction distributed intelligent terminal Beidou navigation technology specifically comprises: Based on the basic coordinate system and the state vector, the relative position and the relative speed of the distributed intelligent terminal are obtained, and an expression is: where: Q z (t) is the state vector of the distributed intelligent terminal relative to the fault at time t, r z (t) = [x z (t),y z (t),Z z (t)] T is the relative position of the distributed intelligent terminal at time t V z (t) = [v x (t),v y (t),v z (t)] T velocity of the distributed intelligent terminal at time t M z (t) = [M G (t), M H (t), M D (t)] is the multi-modal data measured at time t, where M G (t) is the satellite signal, indicating whether the distributed terminal has a fault, taking the value of 1 to indicate that there is a fault, and taking the value of 0 to indicate that there is no fault; M H (t) is the spatial electromagnetic environment data, which affects the signal transmission quality; M D (t) is the terrain information, including contour lines and terrain slope; A dynamics model is constructed, and an expression is: Q z (t+1) = FQ z (t) + Bu(t) + w(t) (14); wherein: Q z (t+1) is the state vector of the distributed intelligent terminal relative to the fault at time t+1, F is the state transition matrix, B is the control matrix, u(t) is the control input, and w(t) is the process noise. A multi-modal data fusion model is constructed, and an expression is: wherein: G(t+1) is the updated state at time t+1, r GNSS (t+1) is the GNSS position, z vision (t+1) is the visual feature at time t+1, and H is the observation matrix. A relative position and speed updating model of fused multi-modal data is constructed, and the relative position and speed of the distributed intelligent terminal are estimated, and an expression of the relative position and speed updating model of fused multi-modal data is: wherein: is a state vector predicted at time t+1, is a relative position and a relative velocity of the intelligent terminal from the intelligent terminal at time t+1, respectively, K(t+1) is a Kalman gain at time t+1, H T is a transpose of an observation matrix, P(t+1|t) is a covariance matrix of a state estimation at time t+1 at time t, P(t|t-1) is a covariance matrix of a state estimation at time t at time t-1, R is an observation noise covariance, and I is an identity matrix having a value of 1 on a diagonal and a value of 0 at other positions.

5. The distributed intelligent terminal fault detection method based on Beidou positioning according to claim 4, characterized in that: The distributed intelligent terminal Beidou information is collected, specifically including: A pseudo-range model between the Beidou satellite and the distributed intelligent terminal is constructed, and an expression is: Where, ρ i Let X be the pseudorange of the i-th Beidou satellite and the distributed intelligent terminal. i ,Y i Z i Let (x0, y0, z0) be the known position of the i-th Beidou satellite in the geocentric-ground-fixed coordinate system, (x0, y0, z0) be the position of the distributed intelligent terminal, c be the speed of light, and δt be the clock offset of the distributed intelligent terminal. i It's a satellite clock deviation; The position (X0, Y0, Z0) of the distributed intelligent terminal and the clock bias δt of the distributed intelligent terminal are solved by using the pseudo-range model between multiple Beidou satellites and the distributed intelligent terminal; A Beidou satellite geographic coordinate conversion model is constructed to convert the position information of the distributed intelligent terminal into geographic coordinates, and an expression of the Beidou satellite geographic coordinate conversion model is: wherein φ is the latitude of the distributed intelligent terminal, λ is the longitude of the distributed intelligent terminal, h is the altitude of the distributed intelligent terminal, R E is the average radius of the Earth.

6. The distributed intelligent terminal fault detection method based on Beidou positioning according to claim 4, characterized in that: The running information of the distributed intelligent terminal is collected, and a distributed intelligent terminal fault detection model is trained based on a Transformer architecture, and the distributed intelligent terminal fault detection model is used for fault detection, specifically including: The running information of the distributed intelligent terminal is collected, and a distributed intelligent terminal fault detection model is trained based on a Transformer architecture, and the distributed intelligent terminal fault detection model is used for fault detection, specifically including: The running information of the distributed intelligent terminal is collected, and a distributed intelligent terminal fault detection model is trained based on a Transformer architecture, and the distributed intelligent terminal fault detection model is used for fault detection, specifically including: X(t) = [T(t), V(t), I(t), P(t)] (3); Wherein X(t) is the running information of the distributed intelligent terminal at time t, T(t) is the temperature collected at time t, V(t) is the voltage collected at time t, I(t) is the current collected at time t, and P(t) is the power collected at time t; The running information of the distributed intelligent terminal is collected, and a distributed intelligent terminal fault detection model is trained based on a Transformer architecture, and the distributed intelligent terminal fault detection model is used for fault detection, specifically including: The expression of the input sample composed of the running information of the distributed intelligent terminal is: Wherein X is the input sample of the Transformer architecture, and X(N) is the running information of the distributed intelligent terminal at time N; X input = X + PE (7); where PE is the position encoding information, t is the time step, g is the dimension index, d is the total dimension of the feature, X input is the input matrix with position encodings added; The position encoding model of the Transformer architecture is constructed, and the expression of the position encoding model of the Transformer architecture is: wherein: Q, K, V are query, key and value vector models, respectively, W Q , W K , W V are weights of the query, key and value vector models, respectively, Attention(Q, K, V) is an attention weight, d k is a dimension of the key vector, K T is a relationship coefficient between the core key and the relevant target in the self-attention mechanism; The self-attention model is constructed, and the mutual relationship between different time steps is calculated, and the expression is: Wherein: MultiHead(Q, K, V) is a multi-head attention feature extraction model, Concat() is a string concatenation function, head j is the attention weight of the jth sample, Q j , K j , V j are the query, key and value vector models of the jth sample respectively; a distributed intelligent terminal fault prediction model based on the Transformer architecture is constructed, and the expression is: y = σ(W out h + b out ) (11) wherein y is the distributed intelligent terminal fault prediction result, h is the feature vector processed through the Transformer architecture, W out and b out are the weights and bias of the output layer, and σ is the activation function. The multi-head attention feature extraction model is constructed, and the expression is: According to the output distributed intelligent terminal fault prediction result y, a threshold τ is set to determine whether there is a fault, and a warning is issued, if the distributed intelligent terminal fault prediction result y is greater than the threshold τ, then output 1, judge the distributed intelligent terminal exists fault, expression is:

Images