A method for calculating position in substation inspection room by integrating QR code recognition and inertial system

Abstract

A method for calculating the position of a substation inspection room that integrates two-dimensional code recognition and an inertial system belongs to the field of computing. A characteristic two-dimensional code tag is set on the switch cabinet, and a camera device is used to collect the characteristic two-dimensional code tag information within the field of view; by relying on the characteristic two-dimensional code tag information within the field of view of the camera device and the inertial data of its own inertial system, the current coordinate position and posture of the mobile robot are comprehensively obtained to solve the problem of accurate positioning of the absolute position of the robot in the substation room. It uses the characteristic two-dimensional code tag set on the switch cabinet to make position corrections, integrates inertial data, and uses a variety of positioning methods according to different positioning information in the environment. Different positioning methods have flexible transitions and complementary error corrections, and have extremely strong anti-interference capabilities. It can be widely used in the design and manufacturing of mobile robots in substation rooms.

Description

Technical Field

[0001] The invention belongs to the field of computing, and in particular relates to an indoor position estimation method for a substation inspection robot. Background Art

[0002] With the development of the national economy, the demand for electricity is increasing, and power supply companies are also deepening the upgrading and transformation of substations. The purpose is to strengthen the inspection and management of the equipment operation status of substations and improve the inspection effect.

[0003] In recent years, a series of mobile robot platforms have been developed at home and abroad for patrol and inspection inside substations (referred to as indoors). There are various modes of movement, including wheeled, tracked, suspended track or roped. With the development of mobile robot platforms, how to accurately determine the position of the robot indoors has become a key node in the development process of mobile robot platforms.

[0004] Satellite positioning technology is a technology that uses satellites to accurately locate objects. It can be used to guide aircraft, ships, vehicles, and individuals to safely and accurately follow the selected route and arrive at the destination on time. However, for indoor environments where satellite signals cannot reach, or for systems that require higher positioning accuracy, an indoor positioning technology that meets the needs of navigation and positioning is still needed. When satellite positioning cannot be used in indoor environments, indoor positioning technology is used as an auxiliary positioning for satellite positioning to solve the problem that satellite signals are weak when reaching the ground, cannot penetrate buildings, and have low positioning accuracy. Ultimately, the current location of the object is located, which is a very important issue in the process of automatic inspection operations indoors in substations.

[0005] In addition to cellular positioning technology based on mobile communication networks, common indoor wireless positioning technologies include Wi-Fi, Bluetooth, infrared, ultra-wideband, RFID, ZigBee, and ultrasound. Due to the limitation of transmission power, such indoor wireless positioning technologies can usually only cover an area with a radius of about 90 meters, and are easily interfered by other signals, thus affecting their accuracy. The energy consumption of the locator is also high.

[0006] Visual understanding technology is an important part of computer vision. It mainly analyzes and understands the content of videos or images with the goal of obtaining information from them. Face detection, face recognition, emotion recognition, text recognition, object classification detection, behavior recognition, scene classification and understanding, and video classification and summarization are all typical visual understanding technologies. In addition, visual understanding can also be used for indoor positioning. By learning the arrangement of indoor objects and identifying object features, a deep understanding of the environment can be achieved, and spatial modeling of the surrounding scenes can be achieved. Then, the relative position of the robot in the room can be obtained, and the robot can use this as a basis for planning its own navigation route.

[0007] This technology currently has a certain amount of technical accumulation, but it has extremely high requirements for artificial intelligence technology, requires a relatively complex and long training process, and has relatively poor stability. If it is to achieve true usability and product level, there is still a long way to go in terms of environmental interpretation, virtual interaction, and environmental construction.

[0008] How to achieve indoor positioning of mobile inspection robots simply, quickly and accurately based on the existing switch cabinet equipment layout of the substation with the help of existing mature technical means is a technical problem that needs to be solved urgently in practical work. Summary of the invention

[0009] The technical problem to be solved by the present invention is to provide a method for calculating the position of a substation inspection room by integrating two-dimensional code recognition and inertial system. It uses the currently very mature two-dimensional code recognition algorithm, calibrates the spatial position of the material two-dimensional code on the switch cabinet, uses this as the absolute position information input, combines the inertial system to calculate the position offset, and corrects the positioning result in real time, thus realizing a very stable and accurate position calculation method for substation indoor switch cabinet inspection.

[0010] The technical solution of the present invention is to provide a method for calculating the position of a substation inspection room by integrating two-dimensional code recognition and an inertial system, including arranging a camera device on a mobile robot used for indoor inspection of the substation, wherein the mobile robot is provided with its own inertial system, and is characterized in that:

[0011] 1) A characteristic two-dimensional code label is set on each switch cabinet in the substation room, and the characteristic two-dimensional code label at least contains the location information of the switch cabinet, and the characteristic two-dimensional code label has uniform size information;

[0012] 2) Using a camera to collect characteristic QR code tag information within the field of view;

[0013] 3) If there is no characteristic QR code tag information within the field of view of the camera device, the last positioning information is used as a reference, and the displacement is calculated based on the inertial data to obtain the absolute position positioning information of the mobile robot and store the information;

[0014] 4) If at least two characteristic two-dimensional code tag information are input within the field of view of the camera device, the absolute position positioning information corresponding to the characteristic two-dimensional code is used to realize the absolute position positioning information of the mobile robot, and the position information is stored;

[0015] 5) When there is only one characteristic QR code tag in the field of view of the camera device, while relying on the characteristic QR code tag information to obtain positioning information, inertial data is also used to assist in calculating the vector displacement from the current time to the last accurate positioning time to obtain the absolute position positioning information of the mobile robot and store the information;

[0016] 6) By relying on the characteristic QR code tag information within the field of view of the camera device and the inertial data of its own inertial system, the current coordinate position and posture of the mobile robot are comprehensively obtained to solve the problem of accurate positioning of the absolute position of the robot indoors in the substation.

[0017] Specifically, the method for calculating the indoor position of the substation inspection room uses the characteristic two-dimensional code tags set on each switch cabinet to make position corrections and integrate inertial data to achieve more accurate indoor position calculation. According to the different positioning information in the environment, a variety of positioning methods are used in combination, and different positioning methods are flexibly transitioned and complementary to each other for error correction, which has a strong anti-interference ability.

[0018] Furthermore, the method for calculating the position in the substation inspection room uses a QR code recognition system to obtain a static image by shooting with a camera device, and uses a Zbar QR code recognition library to complete QR code recognition to obtain the coding content, size, and absolute position of all feature QR codes in the field of view within the camera field of view.

[0019] Specifically, when there are multiple characteristic two-dimensional code tags within the camera's field of view, the position of the mobile robot is reversely determined based on the relative position of each two-dimensional code within the field of view and the absolute position information corresponding to each characteristic two-dimensional code tag.

[0020] Specifically, when there is only one feature QR code tag within the camera's field of view, the position of the mobile robot is calculated based on the absolute position information attached to the feature QR code tag and its relative position within the field of view, and the inertial data of the inertial system is used to assist in calculating the vector displacement from the last QR code feature change to the current time period, so as to avoid accidental errors caused by QR code recognition errors.

[0021] Specifically, when there is no characteristic QR code tag within the camera's field of view, the position of the mobile robot is estimated entirely by relying on the inertial data of the inertial system and the Kalman filter algorithm to ensure relatively accurate indoor positioning within a certain period of time.

[0022] Furthermore, the self-inertial system is composed of a nine-axis high-precision attitude sensor and an optical incremental encoder, wherein the nine-axis high-precision attitude sensor includes a three-axis accelerometer, a three-axis magnetometer and a three-axis gyroscope; the self-inertial system realizes sensor data fusion filtering through a Kalman filter algorithm, and an optical encoder is configured to be connected to the mobile mechanism of the mobile robot, so as to obtain real-time data of the moving speed of the mobile robot.

[0023] The method for calculating the position in a power station inspection room of the present invention uses the positioning information contained in the two-dimensional code tag and the size information of the two-dimensional code tag itself to locate the absolute position of a movable robot in the substation inspection room.

[0024] Compared with the prior art, the advantages of the present invention are:

[0025] 1. The technical solution of the present invention relies on the camera to collect the field of view, and then uses the most mature two-dimensional code recognition tool Zbar as the core of the recognition algorithm. It relies on the two-dimensional code feature information in the field of view and the inertial data of its own inertial system to comprehensively obtain the current coordinate position and posture of the carrier, which can conveniently and quickly solve the problem of precise positioning of indoor robots or devices;

[0026] 2. Apply QR code recognition to the precise positioning of robots in substation rooms, combine it with the inertial system to calculate the position offset, and correct the positioning results in real time, thus improving the accuracy of the position calculation mode for indoor switch cabinet inspections in substations. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a schematic diagram of the framework of the indoor position estimation method of the present invention;

[0028] Figure 2a It is a schematic diagram of reverse positioning of multiple characteristic QR codes in the field of view;

[0029] Figure 2b It is a schematic diagram of a visual field with only one characteristic QR code in the field of view;

[0030] Figure 3 This is a schematic diagram of the signal flow of the attitude estimation and position calculation algorithm based on Kalman filtering. DETAILED DESCRIPTION

[0031] The present invention will be further described below in conjunction with the accompanying drawings.

[0032] The present invention mainly relates to an indoor position estimation method integrating two-dimensional code recognition and inertial system suitable for indoor inspection of substations. The method can comprehensively obtain the current coordinate position and posture of the carrier by relying on the feature information of the two-dimensional code within the field of view and the inertial data of its own inertial system, and can solve the problem of precise positioning of indoor robots or devices.

[0033] The indoor position calculation method involved in the present invention can realize positioning by relying on the absolute position positioning information corresponding to the characteristic two-dimensional code when characteristic two-dimensional code information is input in the field of view environment; when there is no characteristic two-dimensional code information in the field of view, displacement calculation is realized by relying on inertial data based on the last positioning information as a reference, and then the absolute position positioning information is obtained; when there is only one characteristic two-dimensional code in the field of view, the positioning information is obtained by relying on the characteristic two-dimensional code information, and the inertial data is also used to assist in calculating the vector displacement within the time period from the current to the last accurate positioning, so as to avoid accidental positioning errors caused by errors in two-dimensional code recognition; the above three positioning strategies are flexibly connected and transitioned, and have strong anti-interference ability.

[0034] The feature QR code recognition process first relies on the camera to capture the field of view, and then uses the most mature QR code recognition tool Zbar as the core of the recognition algorithm to obtain all the feature QR code data in the field of view, including the QR code encoding content, size and the relative position of the feature QR code in the field of view.

[0035] The position estimation method based on Kalman filtering involved in the present invention can achieve accurate attitude information output and displacement estimation by relying only on inertial data (three-axis accelerometer data, three-axis magnetometer data, three-axis gyroscope data and encoder speed data), and with the absolute positioning information of the previous moment as a reference, it can achieve relatively accurate indoor position estimation within a certain period of time.

[0036] The implementation process of the Kalman filter algorithm used in the indoor positioning calculation process of the present invention is as follows:

[0037] 1) System state equation:

[0038] The state quantity of the integrated navigation Kalman filter is selected as follows:

[0039]

[0040] Among them, δφ e ,δφ n ,δφ u Represent three attitude errors respectively; δv e ,δv n Represent the speed errors in the X and Y directions respectively; δX and δY represent the indoor X coordinate error and indoor Y coordinate error respectively; ε bx , ε by , ε bz Respectively represent the constant drift of the three-axis gyroscope; They represent the zero bias errors of the two-axis accelerometers (X and Y directions).

[0041] The Kalman filter state transfer equation for indoor integrated navigation can be expressed as:

[0042]

[0043] The form of F(t) is as follows:

[0044]

[0045] Among them, F1(t) is the inertial navigation system error matrix, and the non-zero elements are as follows:

[0046] Table 4.7 Non-zero elements in F1(t)

[0047]

[0048]

[0049] F2(t) is the transformation matrix between the basic navigation parameters and the inertial instrument error, and its expression is as follows:

[0050]

[0051] in, is the transformation matrix from the carrier coordinate system to the navigation coordinate system, for The first and second column elements of .

[0052] The system noise transfer matrix G(t) is shown below:

[0053]

[0054] The system noise matrix W(t) is shown below:

[0055] W(t)=[w gx w gx w gx w gx w gx 0 0 0 0 0 0 0] T

[0056] 2) System measurement equation:

[0057] The measurement variables of the integrated navigation Kalman filter are selected as follows:

[0058]

[0059] Among them, v ie -v ge 、v in -v gn , L i -L g , i -λg They respectively represent the X-direction speed, Y-direction speed calculated by inertial navigation, and the difference between X, Y and the corresponding data provided by visual QR code recognition.

[0060] The state transfer matrix H(t) is expressed as follows:

[0061] H(t)=[0 4×3 I a×4 0 4×5 ]

[0062] In summary, the measurement equation of the indoor integrated navigation Kalman filter system can be expressed as:

[0063]

[0064] E(t) is the speed and position error provided by GPS, which are: X-direction speed error, Y-direction speed error, X-coordinate error, and Y-coordinate error.

[0065] Figure 1 The framework of the indoor position estimation method integrates QR code recognition and inertial system. When there is feature QR code information input in the field of view, positioning is achieved by relying on the absolute position positioning information corresponding to the feature QR code; when there is no feature QR code information in the field of view, displacement is estimated by relying on inertial data based on the last positioning information, and then the absolute position positioning information is obtained; when there is only one feature QR code in the field of view, the positioning information is obtained by relying on the feature QR code information, and the inertial data is also used to assist in calculating the vector displacement within the time period from the current to the last accurate positioning, so as to avoid accidental positioning errors caused by QR code recognition errors; the above three positioning strategies are flexibly connected and transitioned, and have strong anti-interference ability.

[0066] Figure 2 is a schematic diagram of the visual reverse positioning of the feature QR code in the field of view. The distance between the feature QR code and the observer (i.e., the aforementioned imaging device or camera) and the relative position offset are reflected in the image as the different sizes of the feature QR code in the field of view. Therefore, by obtaining the size information and position information of the feature QR code, the spatial relationship between the observer and the target QR code can be reversely calculated. Combined with the absolute position information corresponding to the feature QR code, the absolute position information of the observer can be reversely calculated.

[0067] Figure 2a In the scenario where multiple characteristic QR codes appear in the field of view, taking two characteristic QR codes appearing in the field of view as an example, the location information of the two characteristic QR codes is used as input, and the position information corresponding to the center of the field of view can be calculated by interpolation method; Figure 2b The figure in the middle shows a scenario where there is only one characteristic QR code in the field of view. The offset distance between the center of the QR code and the center of the field of view can be determined by the proportional scaling relationship between the size of the QR code and the field of view.

[0068] by Figure 2b For example, if the coordinates of the feature QR code in the field of view are (x3, y3), and the absolute position information of the feature QR code is (X, Y), then the coordinate position corresponding to the center of the observer's field of view is (X-x3, Y-y3).

[0069] Figure 3 This is a signal flow diagram of the attitude estimation and position estimation algorithm based on Kalman filtering. As shown in the figure, the inertial data required for attitude estimation includes three-axis acceleration, three-axis angular velocity and three-axis magnetometer data. The speed data is obtained by an incremental encoder coaxial with the motion platform motor. Combined with the speed data obtained by the inertial system and the accurate value of the last positioning information, the vector displacement of the mobile robot within a certain period of time can be calculated, and then the absolute position information of the mobile robot after this time period can be obtained.

[0070] The technical solution of the present invention is to first set a corresponding QR code on each switch cabinet, and the QR code contains the planar spatial position of the corresponding switch cabinet. The spatial position of the QR code set on the switch cabinet is calibrated and used as the absolute position information input. The field of view is collected by the camera set on the mobile robot, and then the QR code recognition tool Zbar is used as the core of the recognition algorithm to obtain all the characteristic QR code data of the visible switch cabinet in the field of view. The position offset is calculated by combining the inertial data of its own inertial system, and the positioning result is corrected in real time to comprehensively obtain the current coordinate position and posture of the carrier (i.e., the mobile robot). This can solve the problem of precise positioning of indoor robots or devices, and realize extremely stable and accurate robot position calculation for indoor switch cabinet inspection of substations.

[0071] The invention can be widely used in the field of designing and manufacturing indoor movable robots of transformer substations.

Claims

1. A method for calculating the position of a substation inspection room by integrating two-dimensional code recognition and an inertial system, comprising arranging a camera device on a mobile robot used for indoor inspection of the substation, wherein the mobile robot is provided with its own inertial system, and is characterized in that: 1) A characteristic two-dimensional code label is set on each switch cabinet in the substation room, and the characteristic two-dimensional code label at least contains the location information of the switch cabinet, and the characteristic two-dimensional code label has uniform size information; 2) Using a camera device to collect characteristic QR code tag information within the field of view of the camera device; including obtaining a static image by shooting with the camera device, using the Zbar QR code recognition library to complete QR code recognition, so as to obtain the encoding content, size, and position of all characteristic QR codes within the field of view; using the positioning information contained in the QR code tag and the size information of the QR code tag itself to locate the absolute position of the mobile robot in the inspection room of the substation; 3) If there is no characteristic QR code tag information within the field of view of the camera device, the absolute position positioning information of the mobile robot is used as a reference, the inertial data of its own inertial system is relied on, and the Kalman filter algorithm is used to realize the position estimation of the mobile robot, thereby obtaining the absolute position positioning information of the mobile robot and storing the information; 4) If at least two characteristic two-dimensional code label information is input within the field of view of the camera device, the spatial relationship between the camera device and each target two-dimensional code is reversely calculated by obtaining the size information and position information of each characteristic two-dimensional code within the field of view of the camera device, and then the absolute position information of the camera device is reversely calculated based on the absolute position information corresponding to each characteristic two-dimensional code, so as to obtain the absolute position positioning information of the mobile robot, and store the position information; 5) When there is only one characteristic two-dimensional code tag in the field of view of the camera device, while relying on the characteristic two-dimensional code tag information to obtain positioning information, the inertial data is also used to assist in calculating the vector displacement from the current time period to the last accurate positioning time period, so as to obtain the absolute position positioning information of the mobile robot and store the information; specifically including: when there is only one characteristic two-dimensional code tag in the field of view of the camera, the position of the mobile robot is calculated according to the absolute position information attached to the characteristic two-dimensional code tag and its position in the field of view, and the inertial data of the inertial system is used to assist in calculating the vector displacement from the last change of the two-dimensional code feature to the current time period, so as to avoid accidental errors caused by errors in two-dimensional code recognition; 6) By relying on the characteristic QR code tag information within the field of view of the camera device and the inertial data of its own inertial system, the current coordinate position and posture of the mobile robot are comprehensively obtained to solve the problem of accurate positioning of the absolute position of the robot in the substation room; The method for calculating the position of the substation inspection room indoors by integrating the two-dimensional code recognition and the inertial system utilizes the two-dimensional code recognition algorithm to calibrate the spatial position of the characteristic two-dimensional code on the switch cabinet, uses this as the absolute position information input, and combines the inertial system to calculate the position offset, and corrects the positioning result in real time, so as to realize accurate position calculation for the inspection of the indoor switch cabinet of the substation.

2. The method for calculating the position of a substation inspection room by integrating two-dimensional code recognition and inertial system according to claim 1 is characterized in that The self-inertial system is composed of a nine-axis high-precision attitude sensor and an optical incremental encoder, wherein the nine-axis high-precision attitude sensor includes a three-axis accelerometer, a three-axis magnetometer and a three-axis gyroscope; The self-inertial system realizes sensor data fusion filtering through Kalman filtering algorithm, configures an optical encoder to be connected with the mobile mechanism of the mobile robot, and obtains the real-time data of the moving speed of the mobile robot.

Images