A high-precision ground flatness measurement system and method

Abstract

The present invention discloses a high-precision ground flatness measurement system and method. A laser positioning ground station is arranged on the ground, and communicates with an unmanned aerial vehicle flight platform to generate a virtual horizontal plane; the unmanned aerial vehicle flight platform flies in the space above the measured ground, and scans and tests the measured ground point by point; the elevation measurement load is installed on the unmanned aerial vehicle flight platform, and measures the vertical distance to the measured ground downward; the method is to use the unmanned aerial vehicle flight platform and the elevation measurement load, and obtain the ground elevation data in combination with the position of the unmanned aerial vehicle, and obtain the flatness information of the measured ground through data processing and data correction. The present invention performs high-precision measurement of the height of the measured ground, and also measures the uncured cement surface during the construction process, thereby providing reference data for the finishing of the casting surface flatness, supporting the application of automatic brick paving machines, and does not require re-leveling and self-leveling processes, thereby improving the efficiency of building construction.

Description

Technical Field

[0001] The invention belongs to a surface measurement system method in the field of precision measurement technology, and in particular relates to a high-precision ground flatness measurement system and method, which can be used in the field of construction robots. Background Art

[0002] my country is one of the countries with the highest infrastructure levels in the world. In the early stage, it developed rapidly thanks to its advantage in labor costs. Recently, with the increase in labor costs and the demand for construction efficiency, the requirements for infrastructure automation have been increasing, which requires different construction robots and corresponding construction measuring instruments.

[0003] Construction robots have a long history since their first invention, and have gone through three generations from mechanical transmission and hydraulics to robotization. The milestones of the development of construction robots are remote control, automatic and semi-automatic control, and the ability to perform a variety of operations in natural environments. Typical construction robots can be divided into three types according to their common technologies: high-tech operation, high-tech energy-saving, and self-diagnosis of faults. Its research content is rich and its technology coverage is wide. With the development of robot technology, high-reliability and high-efficiency construction robots have entered the market and have broad development and application prospects.

[0004] Construction robots have gone through more than 100 years of history from their invention to the present, and have experienced two generations of mechanical transmission and hydraulic transmission. The current robotized construction machinery is called the third generation, which is a milestone in the development of construction machinery. Construction robots can be remotely controlled, automatically and semi-automatically controlled, and can perform a variety of operations in the natural environment, among which natural operations are the biggest feature. There are many types of construction robots, which can be summarized into three types according to their common technologies: high-tech operation, high-tech energy-saving and self-diagnosis of faults. Its research content is rich and its technical coverage is wide. With the development of robot technology, high-reliability and high-efficiency construction robots have entered the market and have broad development and application prospects. Typical construction robots include construction robots, transportation and loading robots, centralized workstations, etc., which can realize multi-machine collaborative decoration construction operations, realize end-to-end intelligent construction in the decoration stage, build a complete full-cycle construction closed loop, and improve the efficiency and effect of construction.

[0005] During the construction process, the inspection of the paved ground is an important function. By measuring the height of the unsolidified cement ground and comparing it with the horizontal plane, the flatness of the construction ground can be obtained, providing important information for the finishing of the unsolidified ground before solidification.

[0006] During large-scale concrete paving construction, it is necessary to re-inspect the construction surface completed by automated construction robots such as pavers. The current method used is manual single-point measurement, where the surveyor holds a laser elevation detection device and enters the designated position to receive the reference laser and provide elevation data. This existing method of measurement will damage the construction surface, and has low measurement efficiency, poor accuracy, and poor real-time performance of large-area measurement results. It also requires manual participation, which seriously affects the standardization and automation of construction. At present, there is an urgent need for a measurement system that can overcome the above shortcomings and complete the very important process of "real-time, high-precision, and automated re-measurement" in automated paving construction.

[0007] There are two potential non-destructive measurement solutions for the above situation:

[0008] The first method uses a fixed-point high-precision scanning method, that is, a high-precision three-dimensional scanning system is set up at the edge of the ground to be measured to scan the ground. This solution is simple to implement, but it has the disadvantage of a small measurement range. First, the ranging accuracy of the scanner decreases with the increase of distance. The nominal ranging error at 10 meters is greater than ±1mm (faro focus), while the installation height of a general three-dimensional scanning system will not exceed 2 meters from the ground to be measured. In most cases, the installation height is about one person high. The incident angle to the ground at 10 meters is about 80 degrees, which is equivalent to grazing incidence. The ranging error increases rapidly, far exceeding the nominal error. This solution is only suitable for small-scale ground measurement, such as within an incident angle of 60 degrees, that is, within 4 meters around the relative installation point.

[0009] The second method uses a laser tracker to track the flight platform and carry a laser scanner, which effectively solves the distance and incident angle problems of the laser scanner, and always makes the laser scanner range measurement work in a close range (such as 2 meters) and a very small incident angle (almost vertically incident on the ground) mode. The laser tracker of this solution is relatively expensive, and the tracking measurement bandwidth may not be able to eliminate the vibration problem of the flight platform. At the same time, as a laser tracker device, it is impossible to obtain measurement data in real time, and there is no real-time linkage feedback with the flight platform.

[0010] In addition, for construction ground measurement, the traditional technology is to use manpower combined with on-site measurement. This method mainly has several problems:

[0011] The first is immersion measurement, which will destroy the poured ground that has not yet solidified;

[0012] Secondly, there are few measurement sampling points and the speed is slow, so only a limited number of points can be measured;

[0013] The final measurement accuracy is average, which makes it difficult to guide subsequent finishing work and cannot achieve self-leveling accuracy. On this basis, it is impossible to directly carry out tile laying and other work. Summary of the invention

[0014] In view of the demand for high-precision measurement of the flatness of the ground such as paving surfaces during construction, and in order to solve the problems existing in the background technology, the present invention proposes a high-precision ground flatness measurement system and method, which is a solution that uses drone technology in conjunction with a ranging module to achieve high-precision measurement of ground flatness, and fully utilizes the existing drone technology and TOF-based high-precision ranging to form a measurement system, to achieve high-precision non-contact measurement, to achieve high-precision measurement of the ground, and combined with a fitting algorithm, to achieve fast and high-precision measurement of key parameters such as ground flatness, thereby providing a core sensor for ground finishing, which can meet applications in construction and other fields and improve construction efficiency and construction effectiveness.

[0015] At the same time, by adding CCD, linear PSD and other photoelectric detectors to the UAV, the virtual horizontal plane of the laser positioning ground station can be detected to detect the UAV's relative position information as a correction parameter, thereby reducing the accuracy requirements and control bandwidth requirements for the UAV position control.

[0016] The steps of the method of the technical solution adopted by the present invention are as follows:

[0017] 1. A high-precision ground flatness measurement system:

[0018] It includes a laser positioning ground station, which is arranged on the ground and communicates interactively with the UAV flight platform to generate a virtual horizontal plane for controlling the level of the UAV flight platform;

[0019] It includes a UAV flight platform, which flies in the space above the ground to be tested, and scans the ground to be tested and tests the vertical height point by point;

[0020] It includes an elevation measurement load, which is installed on the UAV flight platform and is used to measure the vertical distance to the measured ground downward.

[0021] The altitude measurement load is a high-precision broadband laser ranging module.

[0022] The laser positioning ground station generates a laser beam or a laser surface in the form of laser pulses, and the laser beam or the laser surface constitutes a virtual level surface.

[0023] The laser positioning ground station has a two-dimensional level inside, which provides a horizontal direction and adjusts the laser beam or laser surface to see if it is in a horizontal position. The scanning laser provides a virtual level surface as a reference for height measurement through a beam scanning method.

[0024] The laser positioning ground station is mainly composed of an optical path system and a virtual level control unit.

[0025] The optical path system includes a laser light source, a driving circuit and a beam control optical path connected in sequence, and is used to generate laser pulses within the measurement range;

[0026] The virtual level control unit is mainly composed of two parts: a two-dimensional inclination measurement unit and a leveling actuator. The two-dimensional inclination measurement unit acts as a two-dimensional level to measure the inclination angle of the pulse laser platform and controls the leveling actuator to adjust to a horizontal position to ensure the horizontality of the emitted laser pulse and provide a virtual horizontal plane.

[0027] The elevation measurement payload includes one or a combination of a single-point laser ranging module for single-point ranging and an array laser ranging module for multi-point array ranging based on the time-of-flight measurement principle.

[0028] The single-point laser ranging module has fast measurement speed and high accuracy. It can measure the corners of indoor walls and floors by adding an additional one-dimensional scanning device. The area array laser ranging can solve the problem of identifying the connection between the wall and the ground. The two modules are selected according to the measurement requirements.

[0029] The UAV flight platform is mainly composed of a UAV including a flight control module and a photoelectric detection module; the altitude measurement payload is installed on the UAV, and the UAV carries the altitude measurement payload to complete the relative distance measurement in the vertical direction to the ground; the photoelectric detection module is used to cooperate with the measurement of the virtual horizontal plane formed by the laser pulses emitted by the laser positioning ground station.

[0030] A photoelectric detector is installed on one side of the UAV flight platform.

[0031] The photoelectric detection module measures the virtual horizontal plane formed by the laser pulses emitted by the laser positioning ground station and provides it to the flight control module in the UAV. The flight control module adjusts the attitude of the UAV according to the information measured by the photoelectric detection module to ensure the verticality of the measurement and the height of the UAV itself is within the virtual horizontal plane as a benchmark for elevation measurement.

[0032] 2. A high-precision ground flatness measurement method:

[0033] The measuring method of the present invention utilizes an unmanned aerial vehicle (UAV) flight platform with precise positioning and attitude control, combined with an elevation measurement load installed on the UAV flight platform, to achieve precise measurement of the vertical distance between the UAV and the ground, and obtain ground elevation data in combination with the position of the UAV; the UAV is used to scan and test the ground within the measured range point by point, thereby achieving data distribution of the ground elevation; and through data processing and data correction, the flatness information of the measured ground is obtained, thereby achieving high-precision flatness measurement.

[0034] The method is carried out according to the following steps:

[0035] (1) First, generate a test sequence based on user requirements:

[0036] Generate a test sequence according to the pre-input range area R of the measured ground and the two-dimensional measurement step length;

[0037] (2) Control the UAV to fly to a test point T(i, j) and stabilize it on a virtual horizontal plane using a laser positioning ground station in the height direction;

[0038] (3) Start the height measurement load to measure the vertical distance. Perform l consecutive tests at each test point according to the preset sampling frequency. Each test obtains the vertical distance z(i,j,k), where k represents the test number ordinal number, ranging from 1 to l, and l is the maximum number of test points. All vertical distances z(i,j,k) at each test point obtain the test result sequence z(i,j,k) corresponding to the test point T(i,j);

[0039] (4) Filter the test result sequence z(i,j,k) of each test point T(i,j) to obtain the test result value h(i,j) of the test point T(i,j);

[0040] (5) Repeat steps (2) to (4) until all test points have been tested, and the test result values ​​h(i, j) of all test points T(i, j) form a two-dimensional data array;

[0041] (6) All test result values ​​h(i, j) of the two-dimensional data array are processed to obtain the inclination angles in two orthogonal directions and the actual flatness with the ground.

[0042] The step (1) is specifically as follows: the step length components of the two-dimensional measurement step in the two pre-selected orthogonal directions are dx0 and dy0 respectively; a plurality of test points T(i, j) are arranged in an array with the step length components dx0 and dy0 in the two orthogonal directions within the range area R of the measured ground, T(i, j) represents the test point in the i-th row and j-th column within the range area R of the measured ground, T(i, j) = [x(i), y(j)], wherein x(i) and y(j) respectively represent the coordinates of the test point T(i, j) in the two orthogonal directions, i ranges from 0 to n, j ranges from 0 to m, and n and m respectively represent the total number of test points in the two orthogonal directions within the range area R of the measured ground; and then a test sequence is generated for all the test points in a preset order.

[0043] The filtering in step (4) adopts a l-point average method, which is expressed as:

[0044] h(i,j)=sum(z(i,j,k)) / l

[0045] Among them, sum() represents the sum function, and the average at point l after the sum is the smoothing filter;

[0046] In the step (4), the following outlier removal process is performed before filtering:

[0047] Take the mean and variance of the test result sequence z(i,j,k) of each test point T(i,j), record them as m(k) and std(k) respectively, remove the data that meets the following formula and do not enter the subsequent steps:

[0048] |z(i,j,k)-m(k)|>6×std(k)

[0049] Among them, m(k) and std(k) represent the mean and variance values ​​corresponding to the test result sequence z(i,j,k) of the test point T(i,j), respectively.

[0050] The step (6) is specifically as follows:

[0051] (6.1) According to all the test result values ​​h(i,j) of the two-dimensional data array, a two-dimensional linear fit of the data is performed, and the fitting plane representation formula is:

[0052] h(i,j)=[x h (i), y h (j)]

[0053] f(i,j)=a×x h (i) × dx0 + b × y h (i)×dy0+c

[0054] Among them, x h (i) y h (j) represents the coordinate components of the test result value h(i,j) of the test point T(i,j) in two orthogonal directions, dx0 and dy0 represent the measurement steps on these two orthogonal components, corresponding to the sampling interval, f(i,j) represents the target fitting plane parameters, a, b, c are the first, second, and third fitting quantities respectively;

[0055] The parameter results of the first fitting quantity a, the second fitting quantity b, and the third fitting quantity c are obtained by fitting using the least square method, which are recorded as the first fitting coefficient a0, the second fitting coefficient b0, and the third fitting coefficient c0 respectively;

[0056] (6.2) Then substitute the first fitting coefficient a0, the second fitting coefficient b0 and the third fitting coefficient c0 into the following formula to calculate the fitting parameter f'(i,j) of the test point T(i,j), and then obtain the test result value h(i,j) of all test points T(i,j) and the fitting parameter f'(i,j) to obtain the residual herr(i,j) of the test result value of each test point T(i,j) deviating from the fitting plane:

[0057] f'(i,j)=a0×x h (i) × dx0 + b0 × y h (i)×dy0+c0

[0058] herr(i,j)=h(i,j)-f'(i,j)

[0059] Where f'(i,j) represents the fitting parameter, herr(i,j) represents the residual deviation of the test result value of the test point T(i,j) from the fitting plane;

[0060] (6.3) The first fitting coefficient a0 and the second fitting coefficient b0 are respectively used as the inclination angles of the range area R of the measured ground in two orthogonal directions, and the residual herr(i,j) is used as the actual flatness with the ground, so as to realize the measurement.

[0061] The present invention is a non-contact, non-destructive, rapid measurement system, which is composed of a laser positioning ground station, an unmanned aerial vehicle flight platform, and an elevation measurement load. The laser positioning ground station is used to generate a virtual level surface composed of a laser beam or a laser surface; at the same time, the flight platform is assisted to achieve higher precision positioning, and the unmanned aerial vehicle flight platform completes the measurement path planning, the preliminary flight position positioning, and the scanning measurement of the measured area is realized by the unmanned aerial vehicle flight platform; the elevation measurement load is used, which has the characteristics of high speed and high precision, and measures the elevation difference of the flight platform and the load itself relative to the laser virtual level surface at an extremely high rate, as well as the elevation of the load relative to the ground to be measured, thereby solving the elevation difference of the ground to be measured relative to the laser virtual level surface. At the same time, the high-precision elevation difference real-time data obtained by the load is fed back to the flight control system of the flight platform and the gimbal system, further improving the flight stability of the flight platform.

[0062] The system of the present invention can mark the test results of the test point through the flight path planning unit until the test of all points on the flight route is completed. After error correction, the data of all test points are used through data processing algorithms such as fitting and information extraction to form a full-view topographic map of the ground to be tested and display it. In this way, real-time, high-precision, and automatic retesting of the specified path of the construction surface can be achieved.

[0063] The present invention has the following beneficial effects:

[0064] The system and method of the present invention combine UAV technology and use laser positioning ground reference to provide a virtual plane as its reference. On this basis, distance measurement sensors such as TOF are used to achieve distance measurement, so as to obtain directional information such as the flatness of the measured surface.

[0065] The method of the present invention improves the measurement accuracy, adopts a non-contact measurement method, will not interfere with the construction surface, improves the measurement efficiency, and can provide reference data for the construction and finishing of unsolidified cement surfaces, which is expected to improve the performance indicators of construction such as ground pouring and reduce construction costs; it is expected to be applied to construction robots, and recommends the development of unmanned and automated construction technologies. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 Schematic diagram of the measurement system;

[0067] Figure 2 Schematic diagram of the UAV positioning principle. DETAILED DESCRIPTION

[0068] The present invention will be further described below in conjunction with the accompanying drawings and embodiments:

[0069] like Figure 1 As shown in the figure, the system consists of a laser positioning ground station, a UAV flight platform and an elevation measurement payload.

[0070] In the present invention, non-contact measurement and rapid multi-point measurement are mainly achieved through drones; laser positioning ground stations are used to provide high-precision positioning information for drones and provide a reference platform for measurement; high-precision relative displacement measurement is achieved by using the altitude measurement load carried on the drone flight platform; thus, surface measurement of the measured surface is achieved, and the surface shape information of the measured surface is extracted by methods such as data fitting. This includes three hardware parts and one software part, and the functions of each part are as follows:

[0071] Laser positioning ground station: located outside the measured surface, it uses laser to form a horizontal reference virtual plane (or virtual horizontal plane, virtual horizontal reference plane, etc.), and uses the built-in level meter to make its output laser in a horizontal position, and uses a scanning device or a prism structure to form a virtual horizontal plane. The level meter can achieve an accuracy of 1 micron / millimeter, which corresponds to a horizontal plane accuracy of better than 0.2 arc seconds, which is much higher than the required accuracy; the virtual reference plane formed provides a positioning reference for the drone, and the drone flight platform is maintained within the virtual reference plane through the built-in photoelectric detector.

[0072] The core of the UAV flight platform is the UAV, which provides a non-contact platform; it is equipped with a photoelectric detector to detect the virtual horizontal reference plane provided by the laser positioning ground station, and the UAV is always on this plane through the detection of the laser and the flight control module. Specific detectors such as four-quadrant photoelectric detectors such as QPD can control the position of the UAV by controlling the light spot in the middle of the QPD. At the same time, the UAV is also the implementer of the function of scanning the measurement surface: under the control of the flight control module, according to the pre-planned measurement route and measurement point position, the UAV will fly to the next measurement point after completing one measurement point, and measure the distance after the height is controlled at the virtual horizontal reference plane. The UAV flight platform also has a high-precision high-precision measurement load, which mainly completes the vertical distance measurement. The UAV flight platform can also add other functions, such as data communication between the ground positioning station and the high-speed data channel such as Bluetooth module and wifi wireless module, and uploading the test data in real time.

[0073] The high-precision measurement payload is the core of the measurement, which realizes the measurement of relative distance. When the drone is on the virtual platform, the measurement result represents the height information of the measured point. The high-precision measurement payload can adopt a variety of different technical solutions. In this application, the high-precision measurement is achieved by using the area array TOF (time of flight measurement) as the measurement sensor.

[0074] The main control computer of the system of the present invention realizes the control of the measurement process and the implementation of the algorithm. At the beginning of the measurement, the measurement parameters are input according to the user's needs, including the measurement range, the step length of the measurement points, etc., so as to generate a measurement sequence; then the measurement sequence is sent to the drone as the order of the drone measurement; then the laser positioning ground station is controlled to generate a virtual horizontal reference plane; at the same time, after the high-precision measurement load test is completed or after each point is measured, it communicates with it to complete the acquisition of measurement data; finally, the distance measurement data of each point is obtained by using the measurement data, and then the height data distribution of the entire measured surface is obtained, and finally the data distribution is fitted to obtain the measurement information and parameters such as the slope, flatness, deviation, etc. of the measured surface; the measurement information and parameters are stored, uploaded, and displayed.

[0075] Figure 1 This is a diagram of the measurement system. The ground station in the figure is a laser positioning ground station, which uses the laser output on it to provide a virtual horizontal plane, as shown by the dotted line; the ground station includes a high-precision level, which can adjust the posture of the ground station at the beginning of the measurement based on the level information, thereby ensuring that the accuracy of the virtual horizontal plane formed by the laser plane meets the expected requirements. At present, the accuracy of the high-precision level can reach an accuracy better than 10um / m, which means that within a range of 100 meters, its maximum error will not exceed 1mm. This ensures that the laser virtual horizontal plane can be used as a reference plane.

[0076] The UAV flight platform is also equipped with photoelectric detectors for precise position control. The UAV can be an existing commercial UAV or customized, and it carries the height measurement load; and according to the pre-set measurement route, the height measurement load is distributed and carried and transported to each measurement point, such as Figure 1 As shown in the figure, the drone is suspended in the air and is stably controlled on the laser virtual horizontal plane by the built-in flight control module. Then, the height measurement load measures the relative height of the measured point A on the measured surface. By repeating this process, the data of all pre-set measured points on all measured surfaces can be completed. Based on these data, data fitting can be performed to obtain the surface information of the measured surface.

[0077] The measurement data can be communicated with the ground station through the communication link on the drone and sent to the ground station, which provides data processing functions; optional links include wireless methods such as WiFi and Bluetooth.

[0078] Figure 2 This is a schematic diagram of the positioning principle of UAV. Figure 2 As shown in the figure, the drone is equipped with a photoelectric detector for detecting the height and relative displacement relationship of the virtual horizontal plane formed by the laser. As shown in the figure, it is a typical detector based on a four-quadrant photodiode (QPD), which can also be replaced by four single photodiode (PD) detectors. If the four detection units of the photoelectric detector are respectively recorded as: d1, d2, d3, d4. For the convenience of illustration, the size relationship between the laser horizontal plane and the QPD is enlarged in the figure. The black dotted line corresponds to the virtual laser surface.

[0079] Figure 2 The left and right figures in the figure represent the light signals detected by the QPD when the drone is below and above the horizontal plane, respectively. As can be seen from the figure, if the d1 plus d2 signal is detected to be higher than d3 plus d4, it means that the drone is below the virtual horizontal plane at this time, and the flight control needs to raise its height (left figure); on the other hand, if the d1 plus d2 signal is detected to be lower than d3 plus d4, it means that the drone is above the virtual horizontal plane at this time, and the flight control needs to lower its height; by performing feedback control in this way, knowing that the d1 plus d2 signal is equal to the d1 plus d2 signal, it means that the drone is suspended on the virtual horizontal plane, and its height can be used as a measurement reference. At this time, the height measurement of the measured point vertically below it can be carried out.

[0080] Furthermore, by adjusting the relationship between d1 and d3, and the signal size relationship between d2 and d4, the attitude of the drone can be closed-loop, further improving the measurement accuracy.

[0081] In the specific implementation, the measurement process is carried out according to the following steps:

[0082] (1) First, generate a test sequence based on user requirements:

[0083] Generate a test sequence / test trajectory according to the pre-input range area R of the measured ground and the two-dimensional measurement step length;

[0084] The step length components of the two-dimensional measurement step in two pre-selected orthogonal directions are dx0 and dy0 respectively; multiple test points T(i,j) are arranged in an array with intervals of the step length components dx0 and dy0 in the two orthogonal directions within the range area R of the measured ground, T(i,j) represents the test point in the i-th row and j-th column within the range area R of the measured ground, T(i,j) = [x(i), y(j)], wherein x(i) and y(j) represent the coordinates of the test point T(i,j) in the two orthogonal directions, i ranges from 0 to n, j ranges from 0 to m, and n and m are the total number of test points in the two orthogonal directions within the range area R of the measured ground; then all the test points are generated into a test sequence / test trajectory in a preset order, so as to cover the required measurement range according to the step lengths dx0 and dy0.

[0085] (2) The flight control module controls the drone to fly to a test point T(i, j), and uses the laser positioning ground station to stabilize it on a virtual horizontal plane in the height direction;

[0086] (3) Start the height measurement load to measure the vertical distance. Perform l consecutive tests at each test point according to the preset sampling frequency. Each test obtains the vertical distance z(i,j,k), where k represents the test number ordinal number, ranging from 1 to l, and l is the maximum number of test points. All vertical distances z(i,j,k) at each test point obtain the test result sequence z(i,j,k) corresponding to the test point T(i,j);

[0087] (4) Filter the test result sequence z(i,j,k) of each test point T(i,j) to obtain the test result value h(i,j) of the test point T(i,j);

[0088] The specific filtering adopts the l-point averaging method, which is expressed as:

[0089] h(i,j)=sum(z(i,j,k)) / l

[0090] Before filtering, the following wild point removal process is performed:

[0091] Take the mean and variance of the test result sequence z(i,j,k) of each test point T(i,j), record them as m(k) and std(k) respectively, remove the data that meets the following formula and do not enter the subsequent steps:

[0092] |z(i,j,k)-m(k)|>6×std(k)

[0093] Among them, m(k) and std(k) represent the mean and variance values ​​corresponding to the test result sequence z(i,j,k) of the test point T(i,j), respectively.

[0094] (5) Repeat steps (2) to (4) until all test points have been tested, and the test result values ​​h(i, j) of all test points T(i, j) form a two-dimensional data array of height measurement results;

[0095] (6) All test result values ​​h(i, j) of the two-dimensional data array are processed to obtain the inclination angles in two orthogonal directions and the actual flatness with the ground.

[0096] (6.1) According to all the test result values ​​h(i,j) of the two-dimensional data array, a two-dimensional linear fit of the data is performed, and the fitting plane representation formula is:

[0097] h(i,j)=[x h (i), y h (j)]

[0098] f(i,j)=a×x h (i) × dx0 + b × y h (i)×dy0+c

[0099] The parameter results of the first fitting quantity a, the second fitting quantity b, and the third fitting quantity c are obtained by fitting using the least square method, and are used as the first fitting coefficient a0, the second fitting coefficient b0, and the third fitting coefficient c0, respectively;

[0100] (6.2) Then substitute the first fitting coefficient a0, the second fitting coefficient b0 and the third fitting coefficient c0 into the following formula to calculate the fitting parameter f'(i,j) of the test point T(i,j), and then obtain the test result value h(i,j) of all test points T(i,j) and the fitting parameter f'(i,j) to obtain the residual herr(i,j) of the test result value of each test point T(i,j) deviating from the fitting plane:

[0101] f'(i,j)=a0×x h (i) × dx0 + b0 × y h (i)×dy0+c0

[0102] herr(i,j)=h(i,j)-f'(i,j)

[0103] (6.3) The first fitting coefficient a0 and the second fitting coefficient b0 are respectively used as the inclination angles of the range area R of the measured ground in two pre-selected orthogonal directions, and the residual herr(i,j) is used as the actual flatness with the ground, so as to achieve measurement. The above three types of data are sent to the recording device and displayed through the output device as a signal for ground trimming, etc.

[0104] It can be seen from the implementation that the present invention can measure the height of the measured ground with high precision, and can also measure the uncured cement surface during the construction process, thereby providing reference data for the finishing of the casting surface flatness. The proposed system can become an important part of the building ground casting robot, support the application of automatic brick laying machines, and do not need to re-level and self-leveling processes, thereby improving the efficiency of building construction.

[0105] Figure 2 For the detection of the virtual horizontal plane, photoelectric detectors such as CCD and linear PSD can also be used. As an extension, by measuring the distance between the UAV and the virtual horizontal plane, the control of the UAV's position accuracy can also be reduced. The distance information between the UAV and the virtual horizontal plane can be used to correct the measurement error caused by the UAV position error. At the same time, the measurement bandwidth of CCD, PSD, etc. can be used to improve the measurement speed of the system and reduce the difficulty of UAV control.

Claims

1. A high-precision ground flatness measurement method, characterized in that: The method uses a high-precision ground flatness measurement system: It includes a laser positioning ground station, which is arranged on the ground and communicates interactively with the UAV flight platform to generate a virtual horizontal plane for controlling the level of the UAV flight platform; It includes a UAV flight platform, which flies in the space above the ground to be tested and scans and tests the ground point by point; It includes an elevation measurement load, which is installed on the UAV flight platform and is used to measure the vertical distance to the measured ground downwards; The method is carried out according to the following steps: (1) First, generate a test sequence based on user requirements: Generate a test sequence according to the pre-input range area R of the measured ground and the two-dimensional measurement step length; (2) Control the drone to fly to a test point T(i, j) and stabilize it on a virtual horizontal plane in the height direction using a laser positioning ground station, where i and j represent the i-th row and j-th column in the range area R; (3) Start the height measurement load to measure the vertical distance. Perform l consecutive tests at each test point according to the preset sampling frequency. Each test obtains the vertical distance z(i,j,k), where k represents the test number ordinal number, ranging from 1 to l, and l is the maximum number of test points. All vertical distances z(i,j,k) at each test point obtain the test result sequence z(i,j,k) corresponding to the test point T(i,j); (4) Filter the test result sequence z(i,j,k) of each test point T(i,j) to obtain the test result value h(i,j) of the test point T(i,j); (5) Repeat steps (2) to (4) until all test points have been tested, and the test result values ​​h(i, j) of all test points T(i, j) form a two-dimensional data array; (6) Processing all test result values ​​h(i, j) of the two-dimensional data array to obtain the inclination angles in two orthogonal directions and the actual flatness with the ground; The step (6) is specifically as follows: (6.1) According to all the test result values ​​h(i,j) of the two-dimensional data array, a two-dimensional linear fit of the data is performed, and the fitting plane representation formula is: h(i,j)=[ x h (i),y h (j)] f(i,j)=a×x h (i)×dx0+b×y h (j)×dy0+c Among them, x h (i) y h (j) represents the coordinate components of the test result value h(i,j) of the test point T(i,j) in two orthogonal directions, dx0 and dy0 represent the measurement steps on these two orthogonal components, corresponding to the sampling interval, f(i,j) represents the target fitting plane parameters, a, b, c are the first, second, and third fitting quantities respectively; The parameter results of the first fitting quantity a, the second fitting quantity b, and the third fitting quantity c are obtained by fitting using the least square method, which are recorded as the first fitting coefficient a0, the second fitting coefficient b0, and the third fitting coefficient c0 respectively; (6.2) Then substitute the first fitting coefficient a0, the second fitting coefficient b0 and the third fitting coefficient c0 into the following formula to calculate the fitting parameter f'(i,j) of the test point T(i,j), and then obtain the test result value h(i,j) of all test points T(i,j) and the fitting parameter f'(i,j) to obtain the residual herr(i,j) of the test result value of each test point T(i,j) deviating from the fitting plane: f’(i,j)=a0×x h (i)×dx0+b0×y h (j)×dy0+c0 herr(i,j)=h(i,j)-f'(i,j) Where f'(i,j) represents the fitting parameter, herr(i,j) represents the residual deviation of the test result value of the test point T(i,j) from the fitting plane; (6.3) The first fitting coefficient a0 and the second fitting coefficient b0 are used as the inclination angles of the range area R of the measured ground in two orthogonal directions respectively, and the residual herr(i,j) is used as the actual flatness with the ground, so as to realize the measurement.

2. A high-precision ground flatness measurement method according to claim 1, characterized in that: The step (1) is specifically as follows: the step length components of the two-dimensional measurement step in the two pre-selected orthogonal directions are dx0 and dy0 respectively; a plurality of test points T(i, j) are arranged in an array in two orthogonal directions within the range area R of the measured ground with the step length components dx0 and dy0 respectively, T(i, j) represents the test point in the i-th row and j-th column within the range area R of the measured ground, T(i, j)=[x(i), y(j)], wherein x(i) and y(j) respectively represent the coordinates of the test point T(i, j) in the two orthogonal directions, i ranges from 0 to n, j ranges from 0 to m, and n and m respectively represent the total number of test points in the two orthogonal directions within the range area R of the measured ground; and then a test sequence is generated for all the test points in a preset order.

3. A high-precision ground flatness measurement method according to claim 1, characterized in that: The filtering in step (4) adopts the l-point averaging method, which is expressed as: h(i,j)=sum(z(i,j,k)) / l Among them, sum() represents the sum function, and the average at point l after the sum is the smoothing filter; In the step (4), the following outlier removal process is performed before filtering: Take the mean and variance of the test result sequence z(i,j,k) of each test point T(i,j), record them as m(k) and std(k) respectively, remove the data that meets the following formula and do not proceed to the subsequent steps: |z(i,j,k)-m(k)|>6×std(k) Among them, m(k) and std(k) represent the mean and variance values ​​corresponding to the test result sequence z(i,j,k) of the test point T(i,j), respectively.

4. A high-precision ground flatness measurement method according to claim 1, characterized in that: The laser positioning ground station generates a laser beam or a laser surface in the form of laser pulses, and the laser beam or the laser surface constitutes a virtual level surface.

5. The high-precision ground flatness measurement method according to claim 1, characterized in that: The laser positioning ground station has a two-dimensional level inside, which provides a horizontal direction and adjusts whether the laser beam or laser surface is in a horizontal position.

6. A high-precision ground flatness measurement method according to claim 1, characterized in that: The laser positioning ground station specifically includes an optical path system and a virtual level surface control unit. The optical path system includes a laser light source, a driving circuit and a beam control optical path connected in sequence, and is used to generate laser pulses within the measurement range; The virtual level control unit includes two parts: a two-dimensional inclination measurement unit and a leveling actuator. The two-dimensional inclination measurement unit measures the inclination angle of the pulse laser platform and controls the leveling actuator to adjust to a horizontal position to ensure the horizontality of the emitted laser pulse and provide a virtual horizontal plane.

7. A high-precision ground flatness measurement method according to claim 1, characterized in that: The elevation measurement load includes one or a combination of a single-point laser ranging module for single-point ranging and an array laser ranging module for multi-point array ranging.

8. A high-precision ground flatness measurement method according to claim 1, characterized in that: The UAV flight platform includes a UAV including a flight control module and a photoelectric detection module; an altitude measurement payload is installed on the UAV, and the UAV carries the altitude measurement payload to complete the relative distance measurement in the vertical direction to the ground; the photoelectric detection module is used to cooperate with the measurement of the virtual horizontal plane formed by the laser pulses emitted by the laser positioning ground station.

Images