Inspection robot centralized monitoring system and method

Abstract

The invention discloses an inspection robot centralized monitoring system and method, and relates to the technical field of inspection robot monitoring. The system comprises an inspection robot sensing module, an inspection robot data processing module, an inspection robot intelligent evaluation module and an inspection robot execution module. The inspection robot sensing module is used for acquiring inspection robot data and operation environment data thereof; the inspection robot data processing module is divided into an illumination intensity processing unit for drawing an illumination intensity comparison curve graph, a ground three-dimensional grid model construction unit for constructing a mining area ground three-dimensional grid model, and an electric quantity processing unit for drawing an electric quantity curve graph and constructing a mining area power consumption monitoring model; the inspection robot intelligent evaluation module analyzes the illumination intensity comparison curve graph, the mining area ground three-dimensional grid model and the electric quantity of the inspection robot, and evaluates the abnormal condition of the operation process of the inspection robot; and the inspection robot execution module is used for taking corresponding countermeasures according to the evaluated abnormal conditions.

Description

Technical Field

[0001] The present invention relates to the technical field of patrol robot monitoring, and in particular to a patrol robot centralized monitoring system and method. Background Art

[0002] The background technology of the inspection robot centralized monitoring system mainly stems from the need to improve the traditional manual inspection method. In dangerous or complex environments, manual inspections have many limitations, such as inspection blind spots, low efficiency, and high safety risks. In order to overcome these limitations, the inspection robot centralized monitoring system came into being. The system realizes intelligent monitoring of equipment and environment by integrating sensors, the Internet of Things, data processing and analysis, and remote monitoring technologies. These technologies work together to enable the system to collect, process and analyze inspection data in real time, and promptly discover and deal with potential safety hazards. The application of the inspection robot centralized monitoring system not only improves the efficiency and accuracy of inspection work, but also reduces the cost and risk of manual inspections, providing a strong guarantee for the safe production and intelligent management of enterprises.

[0003] Although the existing centralized monitoring systems and methods for patrol robots have made great progress, there are still some problems that need to be optimized. When the working environment of the patrol robot is insufficiently illuminated, the monitoring image may be blurred and the color may be distorted, reducing the accuracy of fault identification and the overall monitoring effect; secondly, the patrol robots working in mining areas will have worse mobility and increased mechanical wear due to the rugged terrain; and, when the patrol robots are performing their work, there is a lack of real-time power monitoring. When the power is insufficient, the working performance of the patrol robot may be limited, and even the patrol robot may lose power in the middle of the mission, interrupting the patrol mission. Summary of the invention

[0004] In order to solve the above problems, the present invention provides a patrol robot centralized monitoring system, including a patrol robot perception module, a patrol robot data processing module, a patrol robot intelligent evaluation module and a patrol robot execution module;

[0005] The inspection robot perception module is used to obtain inspection robot data and its operating environment data, wherein the inspection robot data is real-time power data, and the operating environment data is light intensity data and three-dimensional point cloud data of the mining area ground;

[0006] The inspection robot data processing module is divided into a light intensity processing unit, a ground three-dimensional grid model construction unit and a power processing unit, wherein the light intensity processing unit is used to set the light intensity threshold and draw a light intensity comparison curve; the ground three-dimensional grid model construction unit is used to build a ground three-dimensional grid model of the mining area; the power processing unit is used to draw a real-time power curve and build a mining area power consumption monitoring model;

[0007] The inspection robot intelligent evaluation module is used to analyze the light intensity comparison curve, the three-dimensional grid model of the mining area ground and the real-time power of the inspection robot to evaluate the abnormal situation of the inspection robot operation process;

[0008] The inspection robot execution module takes corresponding countermeasures according to the assessed abnormal situation.

[0009] Furthermore, the process of acquiring the inspection robot data and its operating environment data includes:

[0010] A1. Install the light sensor on the inspection robot camera to collect the light intensity of its working environment;

[0011] A2. The laser radar emits a laser beam. When the laser beam is projected onto the ground, it interacts with the ground and reflects. The laser radar receives the reflected light signal and calculates the distance between the beam and the ground by combining the displacement formula and the light speed value. In the above process, the laser beam emission angle is changed by the rotating mirror of the laser radar to scan the mining environment in all directions and record the angle of each laser projection point. Combined with the distance and angle of each laser projection point, the laser radar calculates the precise coordinates of the laser projection point in three-dimensional space. Each point contains three coordinate values ​​of x, y and z, thereby obtaining the three-dimensional point cloud data of the mining ground.

[0012] A3. Connect the current sensor to the battery pack of the inspection robot to collect real-time power data.

[0013] Furthermore, in the light intensity processing unit, the process of calculating the light intensity threshold and drawing the light intensity comparison curve diagram includes:

[0014] The light intensity threshold that affects the inspection robot's monitoring image is calculated using the formula:

[0015] L_threshold=L_min+k*(L_max-L_min)

[0016] Among them, L_threshold is the light intensity threshold, L_min is the minimum light intensity, L_max is the maximum light intensity, and k is a coefficient between 0 and 1, which is used to adjust the position of the threshold according to actual needs;

[0017] A two-dimensional coordinate system is established, with the horizontal axis describing time and the vertical axis describing light intensity. The calculated light intensity threshold and real-time light intensity are mapped to the two-dimensional coordinate system with red and black lines respectively to obtain a light intensity comparison curve.

[0018] Furthermore, in the ground three-dimensional grid model construction unit, the process of constructing the ground three-dimensional grid model of the mining area includes:

[0019] B1. Import the 3D point cloud data of the mining area into GIS software for denoising, filtering and registration, and use the tools and algorithms in GIS software to segment and classify the point cloud data and correspond to different ground features;

[0020] B2. Using the Delaunay triangulation algorithm, select a point cloud data as the initial point, construct an initial triangle containing the point, insert other points of the point cloud data into the constructed triangular mesh, find the influencing triangle containing the newly inserted point, delete its common edges, connect the new point and the vertex of the influencing triangle to form a new triangle, locally optimize the newly formed triangle through the minimum internal angle maximum principle and the maximum empty circle principle, and generate a triangular mesh model that describes the ground shape and characteristics of the mining area;

[0021] B3. Calculate the surface normal vector N of the generated triangular mesh model:

[0022] N=AB×AC

[0023] Among them, A(x1, y1, z1), B(x2, y2, z2) and C(x3, y3, z3) form a triangle, AB = (x2-x1, y2-y1, z2-z1), AC = (x3-x1, y3-y1, z3-z1);

[0024] Through the above calculations, the final three-dimensional grid model of the mining area ground is obtained.

[0025] Furthermore, in the power processing unit, the process of setting the light intensity threshold and drawing the light intensity comparison curve diagram includes:

[0026] Set 55% power as the power threshold of the inspection robot, establish a two-dimensional rectangular coordinate system, in which the horizontal axis is time and the vertical axis is the power value, map the real-time power to the coordinate system, and obtain a real-time power curve.

[0027] Furthermore, in the power processing unit, the process of constructing a mining area power consumption monitoring model includes:

[0028] Set the time interval to 10 minutes, calculate the absolute value of the difference between the power consumption in the first 10 minutes and the power consumption in the last 10 minutes, and obtain the power consumption of the inspection robot in the same time period;

[0029] Pre-process the power consumption of the inspection robot in the same time period and the three-dimensional grid model of the ground in the mining area, use the principal component analysis algorithm to combine the three-dimensional grid model of the ground in the mining area into a data matrix, each row describes a sample, each column describes a feature, extract terrain data and surface cover data as feature targets of the three-dimensional grid model of the ground in the mining area, calculate the correlation coefficient of each feature in the data matrix, use the Jacobi method to solve the eigenvalue and eigenvector of the correlation coefficient matrix, analyze the variance contribution rate of each principal component in the data set, and when the contribution rate is greater than 85%, select the principal component as a representative feature, construct each principal component expression, and extract the feature data of the terrain data and the surface cover data, wherein the extracted terrain feature data are elevation, slope and slope direction, and the extracted surface cover feature data are vegetation type, soil type and surface obstacles;

[0030] The random forest parameters are set, and the power consumption of the inspection robot in the same time period, terrain feature data and surface cover feature data are used as data sets. 70% of the data set is divided into a training set and 30% of the data set is divided into a test set. The training set is used to train the initial mining area power consumption detection model. The initial mining area power consumption detection model monitors the power consumption of the inspection robot in the same time period in real time through terrain feature data and surface cover feature data, and obtains the nonlinear relationship between real-time elevation, slope, slope direction, vegetation type, soil type and surface obstacle data and real-time power consumption. The test set is used to evaluate the initial mining area power consumption detection model, and the model parameters are adjusted to optimize the model to obtain the mining area power consumption monitoring model.

[0031] Furthermore, in the intelligent evaluation module of the inspection robot, the process of analyzing the light intensity comparison curve and the three-dimensional grid model of the mining area ground includes:

[0032] Through the light intensity comparison curve, the inspection robot is judged according to the light intensity threshold for normal operation. When the real-time light intensity is lower than the light intensity threshold, the working environment of the inspection robot is dark, which affects the monitoring effect.

[0033] The three-dimensional grid model of the mining area is used to analyze the real-time topography and surface obstacles of the inspection robot's working environment in the mining area. When the inspection robot identifies steep areas such as hills, valleys, and slopes and obstacles in the target movement area, the inspection robot's movement is restricted.

[0034] Furthermore, the process of the inspection robot intelligent evaluation module analyzing the real-time power of the inspection robot includes:

[0035] When the inspection robot is running in a flat area of ​​the working environment, the real-time power curve is analyzed to monitor the power situation in real time; when the inspection robot is running in a mining area, the power consumption monitoring model of the mining area is used to monitor the power situation of the inspection robot in the mining area in real time by using the nonlinear relationship between the real-time elevation, slope, slope direction, vegetation type, soil type and surface obstacle data and the ground characteristics of the mining area and the real-time power consumption;

[0036] When the power level is lower than 55%, the inspection robot's operating performance decreases, and the power level is insufficient to supply the inspection robot to perform a complete task, and the analysis results are wirelessly transmitted to the inspection robot's execution module;

[0037] Furthermore, in the inspection robot execution module, the process of dealing with abnormal situations includes:

[0038] When the light intensity is lower than the light intensity threshold, a command is issued to mobilize the inspection robot to turn on the LED light for supplementary lighting;

[0039] When the inspection robot identifies steep areas such as hills, valleys, and slopes, an execution command is issued to mobilize the inspection robot to switch to the crawler operation mode; when the inspection robot identifies the existence of obstacles in the target movement area, an execution command is issued to mobilize the inspection robot to run around the obstacles;

[0040] When the inspection robot is working in a flat area or in a mining area, if the real-time power level is lower than 55%, the inspection robot's centralized monitoring system will be activated through execution of instructions to send a low-power alarm to the client. The client will receive the instructions and take charging measures to respond.

[0041] Furthermore, a centralized monitoring method for an inspection robot is implemented based on the centralized monitoring system for an inspection robot described above, and is composed of the following steps:

[0042] S1. Use light sensors to collect the light intensity of the working environment of the inspection robot; use laser radar to obtain the three-dimensional point cloud data of the ground in the mining area; use current sensors to collect the real-time power data of the inspection robot;

[0043] S2, the light intensity processing unit calculates the light intensity threshold and draws a light intensity comparison curve; the ground three-dimensional grid model construction unit constructs a ground three-dimensional grid model of the mining area through GIS technology and Delaunay triangulation algorithm; the real-time power processing unit draws a real-time power curve and calculates the power consumption in the same time period; the mining area power consumption monitoring model construction unit uses a random forest algorithm to construct a mining area power consumption monitoring model;

[0044] S3, analyzing the light intensity comparison curve, the three-dimensional grid model of the mining area and the real-time power of the inspection robot, and wirelessly transmitting the analysis results to the inspection robot execution module;

[0045] S4: When the light intensity is lower than the threshold, a command is issued to mobilize the inspection robot to turn on the LED light for supplementary lighting;

[0046] S5. When the inspection robot identifies a steep area such as a hill, a valley, or a slope, an execution command is issued to mobilize the inspection robot to switch to the crawler operation mode; when the inspection robot identifies an obstacle in the target movement area, an execution command is issued to mobilize the inspection robot to bypass the obstacle;

[0047] S6. When the inspection robot is working in a flat area in a mining area, when the real-time power is lower than 55%, the inspection robot centralized monitoring system is mobilized by executing instructions to send a low-power alarm to the client. The client receives the instruction and takes charging measures to respond.

[0048] Beneficial effects of the present invention: Compared with traditional centralized monitoring systems and methods for patrol robots, the centralized monitoring system and method for patrol robots of the present invention closely combine sensor technology, lidar technology, three-dimensional modeling technology, and model building technology with modern information technology, integrate a variety of sensor technologies and lidar technologies, accurately capture the light intensity data and real-time power data of the patrol robot and the three-dimensional point cloud data of the mining area, and achieve real-time and comprehensive monitoring of the centralized monitoring system of the patrol robot through random forest algorithm and big data analysis technology, and obtain light intensity comparison curves after data processing, three-dimensional grid models of the mining area ground, real-time power curves, and mining area power consumption monitoring models, and analyze the light intensity, mining area ground environment, and real-time power consumption of the patrol robot in real time, so as to monitor in real time. Abnormalities can be dealt with in time during the operation of the system, solving the problem that when the working environment of the inspection robot is insufficiently illuminated, the monitoring image is blurred, the accuracy of fault identification and the overall monitoring effect are reduced, the inspection robot working in the mining area has worse mobility due to the rugged ground, the mechanical wear is aggravated, and the inspection robot lacks real-time power monitoring when performing work. The insufficient power may limit the working performance of the inspection robot, and even cause the inspection robot to lose power in the middle of the task, interrupting the inspection task. This ensures that the method in the present invention can refine the dynamic monitoring standard of the centralized monitoring system of the inspection robot within a more precise range, so that the monitored data becomes a more accurate indicator under the same conditions. The development and application of this method have significantly enhanced the degree of intelligence in the working process of the centralized monitoring system of the inspection robot. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 This is a block diagram of a centralized monitoring system for patrol robots;

[0050] Figure 2 The present invention is a flow chart of a centralized monitoring method for inspection robots. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] like Figure 1 and Figure 2 As shown, the present invention provides a technical solution: a patrol robot centralized monitoring system, including a patrol robot perception module, a patrol robot data processing module, a patrol robot intelligent evaluation module and a patrol robot execution module;

[0053] The inspection robot perception module is used to obtain the inspection robot data and its operating environment data, wherein the inspection robot data is real-time power data, and the operating environment data is light intensity data and three-dimensional point cloud data of the mining area ground;

[0054] The inspection robot data processing module is divided into a light intensity processing unit, a ground three-dimensional grid model construction unit and a power processing unit. The light intensity processing unit is used to set the light intensity threshold and draw a light intensity comparison curve chart; the ground three-dimensional grid model construction unit is used to build a ground three-dimensional grid model of the mining area; the power processing unit is used to draw a real-time power curve chart and build a mining area power consumption monitoring model;

[0055] The intelligent evaluation module of the inspection robot is used to analyze the light intensity comparison curve, the three-dimensional grid model of the mining area and the real-time power of the inspection robot, and evaluate the abnormal conditions of the inspection robot operation process;

[0056] The inspection robot execution module takes corresponding countermeasures based on the assessed abnormal situations.

[0057] The process of acquiring inspection robot data and its operating environment data includes:

[0058] A1. Install the light sensor on the inspection robot camera to collect the light intensity of its working environment;

[0059] A2. The laser radar emits a laser beam. When the laser beam is projected onto the ground, it interacts with the ground and reflects. The laser radar receives the reflected light signal and calculates the distance between the beam and the ground by combining the displacement formula and the light speed value. In the above process, the laser beam emission angle is changed by the rotating mirror of the laser radar to scan the mining environment in all directions and record the angle of each laser projection point. Combined with the distance and angle of each laser projection point, the laser radar calculates the precise coordinates of the laser projection point in three-dimensional space. Each point contains three coordinate values ​​of x, y and z, thereby obtaining the three-dimensional point cloud data of the mining ground.

[0060] A3. Connect the current sensor to the battery pack of the inspection robot to collect real-time power data.

[0061] In the light intensity processing unit, the process of calculating the light intensity threshold and drawing the light intensity comparison curve graph includes:

[0062] The light intensity threshold that affects the inspection robot's monitoring image is calculated using the formula:

[0063] L_threshold=L_min+k*(L_max-L_min)

[0064] Among them, L_threshold is the light intensity threshold, L_min is the minimum light intensity, L_max is the maximum light intensity, and k is a coefficient between 0 and 1, which is used to adjust the position of the threshold according to actual needs;

[0065] A two-dimensional coordinate system is established, with the horizontal axis describing time and the vertical axis describing light intensity. The calculated light intensity threshold and real-time light intensity are mapped to the two-dimensional coordinate system with red and black lines respectively to obtain a light intensity comparison curve.

[0066] In the ground three-dimensional grid model construction unit, the process of constructing the ground three-dimensional grid model of the mining area includes:

[0067] B1. Import the 3D point cloud data of the mining area into GIS software for denoising, filtering and registration, and use the tools and algorithms in GIS software to segment and classify the point cloud data and correspond to different ground features;

[0068] B2. Using the Delaunay triangulation algorithm, select a point cloud data as the initial point, construct an initial triangle containing the point, insert other points of the point cloud data into the constructed triangular mesh, find the influencing triangle containing the newly inserted point, delete its common edges, connect the new point and the vertex of the influencing triangle to form a new triangle, locally optimize the newly formed triangle through the minimum internal angle maximum principle and the maximum empty circle principle, and generate a triangular mesh model that describes the ground shape and characteristics of the mining area;

[0069] B3. Calculate the surface normal vector N of the generated triangular mesh model:

[0070] N=AB×AC

[0071] Among them, A(x1, y1, z1), B(x2, y2, z2) and C(x3, y3, z3) form a triangle, AB = (x2-x1, y2-y1, z2-z1), AC = (x3-x1, y3-y1, z3-z1);

[0072] Through the above calculations, the final three-dimensional grid model of the mining area ground is obtained.

[0073] In the power processing unit, the process of setting the light intensity threshold and drawing the light intensity comparison curve includes:

[0074] Set 55% power as the power threshold of the inspection robot, establish a two-dimensional rectangular coordinate system, in which the horizontal axis is time and the vertical axis is the power value, map the real-time power to the coordinate system, and obtain a real-time power curve.

[0075] In the power processing unit, the process of building a mining area power consumption monitoring model includes:

[0076] Set the time interval to 10 minutes, calculate the absolute value of the difference between the power consumption in the first 10 minutes and the power consumption in the last 10 minutes, and obtain the power consumption of the inspection robot in the same time period;

[0077] Pre-process the power consumption of the inspection robot in the same time period and the three-dimensional grid model of the ground in the mining area, use the principal component analysis algorithm to combine the three-dimensional grid model of the ground in the mining area into a data matrix, each row describes a sample, each column describes a feature, extract terrain data and surface cover data as feature targets of the three-dimensional grid model of the ground in the mining area, calculate the correlation coefficient of each feature in the data matrix, use the Jacobi method to solve the eigenvalue and eigenvector of the correlation coefficient matrix, analyze the variance contribution rate of each principal component in the data set, and when the contribution rate is greater than 85%, select the principal component as a representative feature, construct each principal component expression, and extract the feature data of the terrain data and the surface cover data, wherein the extracted terrain feature data are elevation, slope and slope direction, and the extracted surface cover feature data are vegetation type, soil type and surface obstacles;

[0078] The random forest parameters are set, and the power consumption of the inspection robot in the same time period, terrain feature data and surface cover feature data are used as data sets. 70% of the data set is divided into a training set and 30% of the data set is divided into a test set. The training set is used to train the initial mining area power consumption detection model. The initial mining area power consumption detection model monitors the power consumption of the inspection robot in the same time period in real time through terrain feature data and surface cover feature data, and obtains the nonlinear relationship between real-time elevation, slope, slope direction, vegetation type, soil type and surface obstacle data and real-time power consumption. The test set is used to evaluate the initial mining area power consumption detection model, and the model parameters are adjusted to optimize the model to obtain the mining area power consumption monitoring model.

[0079] In the intelligent evaluation module of the inspection robot, the process of analyzing the light intensity comparison curve and the three-dimensional grid model of the mining area ground includes:

[0080] Through the light intensity comparison curve, the inspection robot is judged according to the light intensity threshold for normal operation. When the real-time light intensity is lower than the light intensity threshold, the working environment of the inspection robot is dark, which affects the monitoring effect.

[0081] The three-dimensional grid model of the mining area is used to analyze the real-time topography and surface obstacles of the inspection robot's working environment in the mining area. When the inspection robot identifies steep areas such as hills, valleys, and slopes and obstacles in the target movement area, the inspection robot's movement is restricted.

[0082] The intelligent evaluation module of the inspection robot analyzes the real-time power of the inspection robot, including the following processes:

[0083] When the inspection robot is running in a flat area of ​​the working environment, the real-time power curve is analyzed to monitor the power situation in real time; when the inspection robot is running in a mining area, the power consumption monitoring model of the mining area is used to monitor the power situation of the inspection robot in the mining area in real time by using the nonlinear relationship between the real-time elevation, slope, slope direction, vegetation type, soil type and surface obstacle data and the ground characteristics of the mining area and the real-time power consumption;

[0084] When the power level is lower than 55%, the inspection robot's operating performance decreases, and the power level is insufficient to supply the inspection robot to perform a complete task, and the analysis results are wirelessly transmitted to the inspection robot's execution module;

[0085] In the inspection robot execution module, the process of dealing with abnormal situations includes:

[0086] When the light intensity is lower than the light intensity threshold, a command is issued to mobilize the inspection robot to turn on the LED light for supplementary lighting;

[0087] When the inspection robot identifies steep areas such as hills, valleys, and slopes, an execution command is issued to mobilize the inspection robot to switch to the crawler operation mode; when the inspection robot identifies the existence of obstacles in the target movement area, an execution command is issued to mobilize the inspection robot to run around the obstacles;

[0088] When the inspection robot is working in a flat area or in a mining area, if the real-time power level is lower than 55%, the inspection robot's centralized monitoring system will be activated through execution of instructions to send a low-power alarm to the client. The client will receive the instructions and take charging measures to respond.

[0089] A centralized monitoring method for inspection robots is implemented based on the centralized monitoring system for inspection robots, and consists of the following steps:

[0090] S1. Use light sensors to collect the light intensity of the working environment of the inspection robot; use laser radar to obtain the three-dimensional point cloud data of the ground in the mining area; use current sensors to collect the real-time power data of the inspection robot;

[0091] S2, the light intensity processing unit calculates the light intensity threshold and draws a light intensity comparison curve; the ground three-dimensional grid model construction unit constructs a ground three-dimensional grid model of the mining area through GIS technology and Delaunay triangulation algorithm; the real-time power processing unit draws a real-time power curve and calculates the power consumption in the same time period; the mining area power consumption monitoring model construction unit uses a random forest algorithm to construct a mining area power consumption monitoring model;

[0092] S3, analyzing the light intensity comparison curve, the three-dimensional grid model of the mining area and the real-time power of the inspection robot, and wirelessly transmitting the analysis results to the inspection robot execution module;

[0093] S4: When the light intensity is lower than the threshold, a command is issued to mobilize the inspection robot to turn on the LED light for supplementary lighting;

[0094] S5. When the inspection robot identifies a steep area such as a hill, a valley, or a slope, an execution command is issued to mobilize the inspection robot to switch to the crawler operation mode; when the inspection robot identifies an obstacle in the target movement area, an execution command is issued to mobilize the inspection robot to bypass the obstacle;

[0095] S6. When the inspection robot is working in a flat area in a mining area, when the real-time power is lower than 55%, the inspection robot centralized monitoring system is mobilized by executing instructions to send a low-power alarm to the client. The client receives the instruction and takes charging measures to respond.

[0096] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. Without further restrictions. The sentence "comprises an element defined by ... does not exclude the existence of other identical elements in the process, method, article or device including the element". Although embodiments of the present invention have been shown and described, it is understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the attached claims and their equivalents.

Claims

1. A patrol robot centralized monitoring system, comprising a patrol robot perception module, a patrol robot data processing module, a patrol robot intelligent evaluation module and a patrol robot execution module, characterized in that: The inspection robot perception module is used to obtain inspection robot data and its operating environment data, wherein the inspection robot data is real-time power data, and the operating environment data is light intensity data and three-dimensional point cloud data of the mining area ground; The inspection robot data processing module is divided into a light intensity processing unit, a ground three-dimensional grid model construction unit and a power processing unit, wherein the light intensity processing unit is used to set the light intensity threshold and draw a light intensity comparison curve; the ground three-dimensional grid model construction unit is used to build a ground three-dimensional grid model of the mining area; the power processing unit is used to draw a real-time power curve and build a mining area power consumption monitoring model; The inspection robot intelligent evaluation module is used to analyze the light intensity comparison curve, the three-dimensional grid model of the mining area ground and the real-time power of the inspection robot to evaluate the abnormal situation of the inspection robot operation process; The inspection robot execution module takes corresponding countermeasures according to the assessed abnormal situation.

2. The centralized monitoring system for patrol robots according to claim 1 is characterized in that: The process of acquiring the inspection robot data and its operating environment data includes: A1. Install the light sensor on the inspection robot camera to collect the light intensity of its working environment; A2. The laser radar emits a laser beam. When the laser beam is projected onto the ground, it interacts with the ground and reflects. The laser radar receives the reflected light signal and calculates the distance between the beam and the ground by combining the displacement formula and the light speed value. The rotating mirror of the laser radar changes the laser beam emission angle, scans the mining environment in all directions, and records the angle of each laser projection point. Combining the distance and angle of each laser projection point, the laser radar calculates the precise coordinates of the laser projection point in three-dimensional space. Each point contains three coordinate values ​​of x, y and z, thereby obtaining three-dimensional point cloud data of the mining ground. A3. Connect the current sensor to the battery pack of the inspection robot to collect real-time power data.

3. The centralized monitoring system for patrol robots according to claim 2 is characterized in that: The process of the light intensity processing unit calculating the light intensity threshold and drawing the light intensity comparison curve diagram includes: The light intensity threshold that affects the inspection robot's monitoring image is calculated using the formula: L_threshold=L_min+k*(L_max-L_min) Where L_threshold is the light intensity threshold, L_min is the minimum light intensity, L_max is the maximum light intensity, and k is a coefficient between 0 and 1, which is used to adjust the position of the threshold; A two-dimensional coordinate system is established, with the horizontal axis describing time and the vertical axis describing light intensity. The calculated light intensity threshold and real-time light intensity are mapped to the two-dimensional coordinate system with red and black lines respectively to obtain a light intensity comparison curve.

4. The centralized monitoring system for patrol robots according to claim 3 is characterized in that: In the ground three-dimensional grid model construction unit, the process of constructing the ground three-dimensional grid model of the mining area includes: B1. Import the 3D point cloud data of the mining area into GIS software for denoising, filtering and registration, and use the tools and algorithms in GIS software to segment and classify the point cloud data and correspond to different ground features; B2. Using the Delaunay triangulation algorithm, select a point cloud data as the initial point, construct an initial triangle containing the point, insert other points of the point cloud data into the constructed triangular mesh, find the influencing triangle containing the newly inserted point, delete its common edges, connect the new point and the vertex of the influencing triangle to form a new triangle, locally optimize the newly formed triangle through the minimum internal angle maximum principle and the maximum empty circle principle, and generate a triangular mesh model that describes the ground shape and characteristics of the mining area; B3. Calculate the surface normal vector N of the generated triangular mesh model: N=AB×AC Among them, A(x1, y1, z1), B(x2, y2, z2) and C(x3, y3, z3) form a triangle, AB = (x2-x1, y2-y1, z2-z1), AC = (x3-x1, y3-y1, z3-z1); Through the above calculations, the final three-dimensional grid model of the mining area ground is obtained.

5. The centralized monitoring system for patrol robots according to claim 4 is characterized in that: In the power processing unit, the process of setting the light intensity threshold and drawing the light intensity comparison curve diagram includes: Set 55% power as the power threshold of the inspection robot, establish a two-dimensional rectangular coordinate system, in which the horizontal axis is time and the vertical axis is the power value, map the real-time power to the coordinate system, and obtain a real-time power curve.

6. The centralized monitoring system for patrol robots according to claim 5 is characterized in that: In the power processing unit, the process of constructing a mining area power consumption monitoring model includes: Set the time interval to 10 minutes, calculate the absolute value of the difference between the power consumption in the first 10 minutes and the power consumption in the last 10 minutes, and obtain the power consumption of the inspection robot in the same time period; Pre-process the power consumption of the inspection robot in the same time period and the three-dimensional grid model of the ground in the mining area, use the principal component analysis algorithm to combine the three-dimensional grid model of the ground in the mining area into a data matrix, each row describes a sample, each column describes a feature, extract terrain data and surface cover data as feature targets of the three-dimensional grid model of the ground in the mining area, calculate the correlation coefficient of each feature in the data matrix, use the Jacobi method to solve the eigenvalue and eigenvector of the correlation coefficient matrix, analyze the variance contribution rate of each principal component in the data set, and when the contribution rate is greater than 85%, select the principal component as a representative feature, construct each principal component expression, and extract the feature data of the terrain data and the surface cover data, wherein the extracted terrain feature data are elevation, slope and slope direction, and the extracted surface cover feature data are vegetation type, soil type and surface obstacles; The random forest parameters are set, and the power consumption of the inspection robot in the same time period, terrain feature data and surface cover feature data are used as data sets. 70% of the data set is divided into a training set and 30% of the data set is divided into a test set. The training set is used to train the initial mining area power consumption monitoring model. The initial mining area power consumption monitoring model uses terrain feature data and surface cover feature data to monitor the power consumption of the inspection robot in the same time period in real time, and obtains the nonlinear relationship between real-time elevation, slope, slope direction, vegetation type, soil type and surface obstacle data and real-time power consumption. The test set is used to evaluate the initial mining area power consumption monitoring model, and the model parameters are adjusted to optimize the model to obtain the mining area power consumption monitoring model.

7. The centralized monitoring system for patrol robots according to claim 6 is characterized in that: In the inspection robot intelligent evaluation module, the process of analyzing the light intensity comparison curve and the three-dimensional grid model of the mining area ground includes: Through the light intensity comparison curve, the inspection robot is judged according to the light intensity threshold for normal operation. When the real-time light intensity is lower than the light intensity threshold, the working environment of the inspection robot is dark, which affects the monitoring effect. The three-dimensional grid model of the mining area is used to analyze the real-time topography and surface obstacles of the inspection robot's working environment in the mining area. When the inspection robot identifies steep areas such as hills, valleys, and slopes and obstacles in the target movement area, the inspection robot's movement is restricted.

8. The centralized monitoring system for patrol robots according to claim 7 is characterized in that: In the inspection robot intelligent evaluation module, the process of analyzing the real-time power of the inspection robot includes: When the inspection robot is running in a flat area of ​​the working environment, the real-time power curve is analyzed to monitor the power situation in real time; when the inspection robot is running in a mining area, the power consumption monitoring model of the mining area is used to monitor the power situation of the inspection robot in the mining area in real time by using the nonlinear relationship between the real-time elevation, slope, slope direction, vegetation type, soil type and surface obstacle data and the ground characteristics of the mining area and the real-time power consumption; When the power level is lower than 55%, the inspection robot's operating performance decreases, and the power level is insufficient for the inspection robot to perform a complete task, and the analysis results are wirelessly transmitted to the inspection robot's execution module.

9. The centralized monitoring system for patrol robots according to claim 8, characterized in that: In the inspection robot execution module, the process of dealing with abnormal situations includes: When the light intensity is lower than the light intensity threshold, a command is issued to mobilize the inspection robot to turn on the LED light for supplementary lighting; When the inspection robot identifies steep areas such as hills, valleys, and slopes, an execution command is issued to mobilize the inspection robot to switch to the crawler operation mode; when the inspection robot identifies the existence of obstacles in the target movement area, an execution command is issued to mobilize the inspection robot to run around the obstacles; When the inspection robot is working in a flat area or in a mining area, if the real-time power level is lower than 55%, the inspection robot's centralized monitoring system will be activated through execution of instructions to send a low-power alarm to the client. The client will receive the instructions and take charging measures to respond.

10. A method for centralized monitoring of inspection robots, implemented based on the centralized monitoring system for inspection robots described in any one of claims 1 to 9, characterized in that: It consists of the following steps: S1. Use light sensors to collect the light intensity of the working environment of the inspection robot; use laser radar to obtain the three-dimensional point cloud data of the ground in the mining area; use current sensors to collect the real-time power data of the inspection robot; S2, the light intensity processing unit calculates the light intensity threshold and draws a light intensity comparison curve; the ground three-dimensional grid model construction unit constructs a ground three-dimensional grid model of the mining area through GIS technology and Delaunay triangulation algorithm; the real-time power processing unit draws a real-time power curve and calculates the power consumption in the same time period; the mining area power consumption monitoring model construction unit uses a random forest algorithm to construct a mining area power consumption monitoring model; S3, analyzing the light intensity comparison curve, the three-dimensional grid model of the mining area and the real-time power of the inspection robot, and wirelessly transmitting the analysis results to the inspection robot execution module; S4: When the light intensity is lower than the threshold, a command is issued to mobilize the inspection robot to turn on the LED light for supplementary lighting; S5. When the inspection robot identifies a steep area such as a hill, a valley, or a slope, an execution command is issued to mobilize the inspection robot to switch to the crawler operation mode; when the inspection robot identifies an obstacle in the target movement area, an execution command is issued to mobilize the inspection robot to bypass the obstacle; S6. When the inspection robot is working in a flat area in a mining area, when the real-time power is lower than 55%, the inspection robot centralized monitoring system is mobilized by executing instructions to send a low-power alarm to the client. The client receives the instruction and takes charging measures to respond.

Images