Application of indoor and outdoor integrated collaborative positioning technology in power transportation detection

With the continuous acceleration of the construction of electric power Internet of Things, the structure and scale of electric power system is more massive, the complexity and danger coefficient of field operation is enhanced, and the risk of causing safety accidents is further increased. Therefore, based on wireless sensor network, collecting and transmitting information of personnel and site operation environment, establishing power operation tracking system is of great significance to improve the safety and efficiency of operation and inspection personnel. At present, most of the electric power operation and maintenance personnel are positioned by satellite or wireless WIFI, which has a large error or even cannot be positioned, so this paper proposes a 5G+Beidou fusion indoor-outdoor cooperative positioning technology for electric power operation and inspection service. The fusion model of Beidou carrier phase differential positioning and 5G cellular network positioning is established by using the improved Federal Kalman filter algorithm, which can avoid data pollution caused by Beidou receiver failure or 5G communication module failure in a complex environment, and avoid the inability to locate compared with the common Federal Kalman filter. phenomenon.


Introduction
With the complexity and danger of power operations, which are prone to safety accidents, and with the increasing workload of power operation and maintenance, it is important to establish a power operation tracking system based on wireless sensor networks [1][2][3][4][5] to collect information about operation and inspection personnel and the field operation environment to improve personnel safety and work efficiency [6].
At present, most of the electric power operation and inspection personnel are positioned by positioning sensing methods such as satellite or wireless WIFI, and literature [7] proposed a power grid intelligent inspection and maintenance system based on BeiDou satellite technology, which realized automatic personnel positioning, wireless data return and background query management.In response to the current problems of low automation of manually operated UAVs for transmission line patrol and difficulties in standardization and standardization of manually operated UAVs, resulting in long patrol time or multiple crashes, literature [8] proposed a multi-rotor UAV intelligent patrol method for transmission lines based on carrier phase differential positioning technology around the goal of unmanned autonomous patrol by UAVs proposed by the Southern Power Grid Company.The literature [9] proposed a Beidou and UWB-based positioning system for power inspectors to solve the safety problem of power inspectors.The literature [10] proposes a distribution network fault location and isolation method based on the integrated energy Internet, and designs a model and algorithm applicable to distribution network fault location and isolation in the integrated energy Internet, which achieves the goal of efficient and fast location, while reducing the computation and storage volume.To further enhance the control capability of military equipment in military logistics warehousing and improve the role of military logistics efficiency and quality, literature [11] proposed a 5G and high-precision positioning fusion in military logistics warehousing scheme.The literature [12] studied a multi-point positioning algorithm based on base stations to calculate the longitude and latitude location information of mobile terminal users and proposed a raster-level base station network coverage analysis method.In the paper [13], a study was conducted on 5G IPRAN associated base station fast location network, and a simple network management protocol was used to bridge the gap between wireless and transmission information to achieve fast location of network problems, reduce the difficulty of network operation and maintenance, and realize the simplification of operation and maintenance.
While wireless positioning technology can effectively improve the efficiency of power system operation and maintenance and reduce the risk of grid failure.To the best of our knowledge most of the research about the electric power transportation inspection personnel are positioned through a single positioning method such as satellite or wireless WIFI, which has large errors, and even cannot be located in a closed space environment such as indoors.
The main contributions are summarized as follows: The 5G+Beidou integrated indoor and outdoor coordinated positioning technology for power transportation inspection services is proposed.Relying on the high density of 5G base stations, strong indoor and outdoor coverage, and large communication bandwidth, it can effectively solve the lack of traditional satellite positioning coverage and urban canyon positioning accuracy.Low-level issues, An improved federated Kalman filter algorithm is proposed to establish a fusion model of Beidou carrier phase differential positioning and 5G cellular network positioning.Compared with the commonly used federated Kalman filter, it can avoid data caused by Beidou receiver failure or 5G communication module failure Pollution problems in a complex environment, the best effect can reach decimeter level, to avoid the phenomenon of unable to locate.
The simulation results show that verify the effectiveness of the 5G+Beidou fusion positioning technology proposed in this paper.
to the electric transportation inspection personnel, equipment and vehicles is obtained as ( , , )  X Y Z , the specific calculation formula is Where: c is the speed of light, t  is the clock error of the receiver.When the power grid maintenance personnel, equipment and vehicles are operating in the suburbs, there is less electromagnetic signal interference.The positioning errors mainly include satellite ephemeris and clock errors, ionospheric and tropospheric errors during signal propagation, and errors caused by delays, and receiver observation noise , Clock and delay errors, etc., but due to the wide field of view, there are more satellites available, which greatly reduces the positioning error, and the actual application effect is good; but when the power grid operation and maintenance personnel, equipment and vehicles are densely populated in urban areas and substations In confined space environments such as power distribution rooms, not only the electromagnetic interference is strengthened, but the large number of obstructions around greatly reduces the available satellites.The positioning error is too large, which makes the positioning data unusable.In actual use, it often shows positioning drift and no positioning.It causes inconvenience to the real-time tracking and maintenance of electric power transportation and inspection personnel, equipment and operation and maintenance vehicles.

System model
As shown in Figure 1, domestic operators are currently promoting the joint deployment of 5G base stations and RTK reference base stations.While reducing operation and maintenance costs, they use the large broadband and low latency characteristics of 5G networks to complete the transmission of carrier phase observation information.Broadcasting and solving information, realizing fast and accurate positioning.At the same time, 5G uses its own base station positioning technology to obtain the locations of indoor substations, power distribution rooms, and inspection personnel, and send them to the corresponding equipment.The fusion positioning equipment completes the fusion processing of Beidou positioning information and 5G cellular network positioning information.The RTK technology determines the three-dimensional coordinates of the observation point by processing the carrier phase of the two observation stations in real time.The realization method can be divided into the correction method and the difference method: the former is to calculate the correction amount by referring to the position of the base station and the satellite ephemeris information.The coordinates are sent to the user receiver to calculate the coordinates, which greatly reduces the calculation burden of the user receiver, and can achieve decimeter-level positioning accuracy; the latter uses the carrier phase of the known reference station and the carrier phase of the receiver to be measured.The coordinates can be calculated by difference, which can achieve centimeter-level or even millimeterlevel high precision.In this paper, a correction method is adopted based on the performance of the user receiver.The principle of carrier phase difference is shown in Figure 2.
Then the pseudorange observation value between the satellite and the reference station can be obtained from the carrier wavelength and the phase difference of the satellite station Where  is the wavelength, 0 ( ) N t is the initial period ambiguity; 0 ( ) is the whole cycle count;  is the fractional part of the observed phase.Knowing j i  , we can get Where: i t  is the clock error of the reference station, j t  denotes the satellite clock error, We can obtain the pseudorange correction number at the base station as Transmit this correction number to the user station for correction, which is expressed by When the base station and the 5G communication base station are jointly deployed, based on the density of the 5G base station, it can be considered j Where ( ) ( ) ( ) , if the satellite tracking does not lose the lock during the re-observation process, 0 ( ) j N t is a constant.Let , then the formula (8) can be expressed as Where, 0 ( ) Z are all constants, due to the difference between the clock difference of the two receivers, the noise difference and the multipath effect between the two stations, the variation between adjacent epochs is less than the allowable error,   can also be regarded as a constant.Therefore, by observing 4 and more satellites at the same time, the locomotive can be positioned.

5G cellular network positioning
The 5G cellular network adopts high-frequency millimeter wave communication, supports multibandwidth reception and transmission, and propagates almost in a straight path with high resolution, which makes it easy to achieve higher precision ranging and angle measurement.In a 5G cellular network, according to the characteristics of wireless communication technology, it is easier to measure 0 0 [ ( ) ( )] ( ) the radio wave propagation time, distance, angle, and signal strength between the target's transmission signal and the base station, and then combine the relative position information between different base stations to calculate Go out of the target position to achieve positioning.The positioning methods based on 5G base stations mainly include time of arrival (OAL), time difference of arrival (TDOA), angle of arrival (AOA), and Received Signal Strength (RSS) and other methods, in order to improve positioning accuracy, can use TDOA/AOA combined positioning methods.Assuming that the coordinates of M base stations is ( , , ) x y z , the user coordinates is ( , , )  x y z , each base station can obtain the azimuth angle  , the elevation angle  and the observation  of the TDOA of other base stations except the main base station, then there is arctan( ), 1, 2,..., arctan( ), 1,2,..., ( ) ( ) ) , 1 ,2,..., V denote the observation noise, the model of user positioning can be expressed as , , .., ,..., ,..., , , ,..., ,..., ,..., TDOA/AOA combined positioning requires accurate measurement of the radio wave transmission time and angle.The main factors affecting the error are time measurement accuracy, angle quantization, beam width, distance and so on.

Fusion model based on improved federated Kalman algorithm
Problem, this article adds local filter fault self-checking and input control modules on the basis of the commonly used federated Kalman filter model, and designed an improved federated Kalman filter model as shown in Figure 3: Output the position vector ˆi X and the covariance matrix i P to the main filter for optimal fusion of the data, and then output the optimal values g X and g P .At the same time, the optimal value and information distribution factor

Beidou positioning local filter design
The Beidou positioning local filter is designed in a two-dimensional space to define the state vector ( ), ( ), ( ), ( ) Organize available Where The state equation of the subsystem is In the same way, assuming that the coordinate of the object at time k is ( ( ), ( )) z z x k y k , its coordinate equation can be expressed as The observation equation of the available system is Where,

Design of local filter for 5G cellular network positioning
The corresponding channel model is NLOS, and the signal quality will be affected by obstructions.Suppose ( , , ) x y z is the position coordinate of user p at the moment, and the instantaneous speed is , , , ( , , ) . Using the positioning method that combines TOA and AOA, there are distance and angle measurement errors, which affect the authenticity of the data.In order to improve accuracy, Kalman filtering can be used for data processing, which can be described as Where, f is determined by the target motion law; function h is the TDOA/AOA combined positioning model in the solution.The state vector at user p can be expressed as x y z v v v  S ;  is the state transition matrix; the covariance matrix of the noise vector k W is Q , then the state equation of the user at time p can be expressed as 1 0 0 0 0 0 1 0 0 0 0 0 1 0 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  is the variance of the driving noise, which is used to make up for the limitation of the matrix  describing the law of motion, and it is suitable for any law of motion.The observation equation of the target is defined as follows ( ) Where: p Z is the observation data vector; p V is the observation noise.Since ( ) p h S is a nonlinear transformation, in order to apply the extended Kalman filter, the linearized observation equation is as follows The Kalman filter can dynamically restore the true positioning information of the target in real time and remove the observed noise and interference.Whenever new observation data is acquired, combined with the predicted value at the previous moment, the best estimate of the current location can be obtained.

Main filter data fusion
In order to ensure the final fusion accuracy of the main filter, the local filter must first self-check itself to determine whether the subsystem is malfunctioning, and decide whether to input data to the main filter.

 Fault self-detection
There is no communication base station or communication card.5G location failure caused by costs, communication card failure, communication module damage, etc.When the subsystem fails, the state output data of the corresponding local filter changes greatly, so the ordinary residual check method can be used for detection, and the calculation formula is as follows Where  is the residual, Z is the measured value,   1 Z k  is the observed value.When  is greater than the set value, it can be considered that the sub-filter has failed and stop inputting data to the main filter  Data Fusion After the local filter processes the positioning information, the main filter completes the fusion of the information and outputs the optimal value.The fusion algorithm is as follows The optimal result is fed back to the local filter as Where: P represents the covariance matrix; Q denotes the noise covariance matrix; X denotes the state vector; i  represents the information allocation factor, which satisfies Due to the complex indoor environment, the use of fixed information distribution factors will greatly affect the accuracy of the fusion data.Considering that there will be a large number of 5G base stations in the city in the future, the signal will be relatively stable, and the information distribution factor can be dynamically changed according to the number of available satellites.The degree of influence of occlusion on positioning is mainly divided into three situations: (1) There is no occlusion, the positioning is normal, and the distribution factors of the two are equal.
(2) There is occlusion, the positioning effect is poor, and the allocation factor of Beidou becomes smaller as the number of available satellites decreases.
(3) Completely occluded and unable to locate, the distribution factor of Beidou approaches zero.The relationship between the number of available satellites and the value of the information.distribution factor is shown in Table 1.

Simulation
In order to verify the effectiveness of the method, we choose to simulate the relatively open mountainous transmission line environment, the complex transmission environment in the city, and the positioning effect in different environments such as indoor substations and power distribution rooms.That is: in an environment where the number of 5G base stations is stable, the positioning effect of the fusion of Beidou and 5G is verified according to the number of available Beidou satellites, and at the same time it is compared with Beidou's separate positioning.The positioning error can be divided into horizontal and vertical directions according to the three-dimensional coordinates, as shown in Figure 4, Figure 5 and Figure 6.The error in the vertical direction decreases as the number of available Beidou stars increases.As shown in Figure 4, the two methods of single Beidou positioning and fusion positioning are compared in different environments.When simulating the relatively empty mountainous transmission line environment in the city, that is, when the Beidou available stars are 6 or more, the minimum error of the former is about 3.7m, while the minimum error of the latter is only 2m; when the complex environment in the city is simulated, the Beidou available stars When it is less than 6 but more than 2, the error of the latter is about 2m lower than that of the former.As shown in Figure 6, when simulating indoor environments such as substations and power distribution rooms, that is, when the available Beidou stars are less than 3, the former fails to locate, and the latter has a slightly higher error, but the location service is still available, at about 4m.In order to simulate indoor substations, power distribution rooms and other signal blocking scenarios, the number of observable base stations in the experiment is set to a uniform distribution between 1 and 2. Due to the small number of observable 5G base stations, TOA-based 5G positioning cannot be performed.The addition of AOA observations reduces the need for the number of observable 5G base stations in the positioning calculation, so that the 5G positioning based on TOA/AOA can be calculated to obtain positioning results and the percentage of successful epochs has reached 98%, as shown in Table 2.The addition of AOA observation when the base station signal is blocked has achieved a high success rate of positioning, but the positioning accuracy is still reduced compared to the situation where the number of base stations is sufficient.After further adding inertial sensors, the position error is reduced by about 70%, that is, the TOA/AOA integrated navigation can achieve 100% positioning success rate while eliminating flying spots with large errors.The accuracy comparison of positioning results is shown in Table 3 Table 3 It can be seen from the Table3 the Beidou and 5G fusion positioning solution in different environments has significantly higher positioning accuracy than the Beidou technology alone, and the best effect can reach the decimeter level.Using the improved federated Kalman algorithm for centralized filtering can give the best error estimate.When the Beidou available satellites are insufficient, the information distribution factor of increasing 5G positioning is adopted to reduce the disadvantage of less Beidou available satellites, which can effectively improve the positioning accuracy and meet the needs of power system inspection positioning in complex environments.

Conclusions
In this paper, a new type of 5G+Beidou integrated indoor and outdoor collaborative positioning technology for electric power transportation inspection services is proposed.Specifically, First, the principles of 5G positioning and Beidou positioning technology are studied, and then based on this, using an improved federated Kalman filter algorithm, a positioning method based on the fusion of Beidou carrier phase difference and 5G cellular network technology is proposed to eliminate and reduce the presence in the positioning process.The noise and error.The simulation results show that the fusion

Figure 1 .
Figure 1.RTK reference base station and 5G base station integration model.

1 k
operators, the eastward position vector and the speed vector of the equipment, and ( ) operators, the northward position vector and the equipment Speed vector, T is the time interval; ( ) independent orthogonal vectors of Gaussian white noise k V ( ) with zero mean and variance 2 v  .According to the motion formula of personnel and equipment, operation and maintenance of on-site sports personnel and the equation for device  at time as follows.

Figure 4 .
Figure 4. Comparison of vertical error data.

Figure 5 .
Figure 5.Comparison of horizontal error data.

Figure 6 .
Figure 6.Relationship between transmitting power and ranging accuracy.

1 i
  are fed back to the local filter.This model can avoid data pollution problems caused by Beidou receiver failure or 5G communication module failure, and effectively improve positioning accuracy.

Table 1 .
Information distribution factor values in different environments.

Table 2 .
Position error statistics of different 5G positioning algorithms when the base station is blocked.
. Accuracy comparison of positioning results.