Distributed photovoltaic robots and their usage methods

Abstract

This invention proposes a distributed photovoltaic robot and its usage method, belonging to the technical field of photovoltaic panel cleaning equipment. It includes two parallel and spaced tracks; a main unit straddling the two tracks and slidable relative to them; a water tank mounted on the main unit for supplying water for cleaning the photovoltaic panels; a cleaning trolley unit positioned on the surface of the photovoltaic panel on one side of the main unit and connected to the movable end of a hoisting mechanism, used for cleaning photovoltaic panels near the main unit; and a hoisting mechanism located at the end of the main unit furthest from the tracks and connected to the cleaning trolley unit, used to guide water from the water tank to the cleaning trolley unit and change the distance between the cleaning trolley unit and the main unit. This invention has good applicability, can efficiently plan the cleaning trajectory on the photovoltaic panels, and offers higher safety.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic panel cleaning equipment technology, and in particular to a distributed photovoltaic robot and its usage method. Background Technology

[0002] In recent years, with the rapid development of renewable energy, photovoltaic power plants have become an important means of power supply. Most of the large-scale ground-mounted photovoltaic power plants already built in China are tens of megawatts or more in scale. Distributed photovoltaic power plants, however, are smaller in scale and more dispersed, making photovoltaic cleaning a large-scale and challenging task. Conventional manual cleaning methods require significant manpower and time costs, and the cleaning work also carries certain risks. Regular cleaning is necessary to ensure the continuous and stable operation of distributed photovoltaic power plants. Although photovoltaic cleaning robots have made some progress in certain aspects, they still have some shortcomings and deficiencies: 1) Existing photovoltaic cleaning robots lack adaptability and flexibility to diverse scenarios and cannot effectively handle photovoltaic panels of different types and shapes; 2) Photovoltaic cleaning robots face limitations in efficiency and speed when cleaning photovoltaic panels, with most robots cleaning slowly.

[0003] Given the characteristics of traditional distributed power stations—wide distribution, rapid growth, and urgent need for maintenance—it is essential to develop an intelligent cleaning robot and its application method for distributed photovoltaic power stations in high-altitude environments, providing higher safety, consistency in cleaning quality, and high efficiency in cleaning work. Summary of the Invention

[0004] In view of this, the present invention proposes a distributed photovoltaic robot with good adaptability, wide applicability, and high cleaning efficiency, as well as a method of using it.

[0005] The technical solution of this invention is implemented as follows:

[0006] On one hand, the present invention provides a distributed photovoltaic robot, comprising:

[0007] Two tracks, parallel to each other and spaced apart;

[0008] The main unit is straddling the two tracks, and moves linearly relative to the extending direction of the two tracks.

[0009] A water tank, installed on the main unit, is used to provide water for cleaning the photovoltaic panels;

[0010] A cleaning trolley unit is installed on the surface of the photovoltaic panel on one side of the main unit. The cleaning trolley unit is used to clean the photovoltaic panel near the main unit.

[0011] The hoisting mechanism is located at the end of the main unit away from the track and is connected to the cleaning trolley unit; it is used to guide water from the water tank to the cleaning trolley unit and change the distance between the cleaning trolley unit and the main unit.

[0012] Based on the above technical solutions, preferably, the main unit includes a load-bearing plate, a load-bearing beam, and a GPS positioning device; the load-bearing plate is disposed between two tracks and extends outward in a direction away from the two tracks, and a water tank is disposed at the end of the load-bearing plate away from the ground; two traveling mechanisms are symmetrically disposed on both sides of the load-bearing plate along the axial extension direction of the tracks, the two traveling mechanisms are fixedly connected to the load-bearing plate, and the movable ends of the two traveling mechanisms are rotatably connected to the tracks; the load-bearing beam is disposed at the end of the load-bearing plate away from the two tracks, and a hoisting mechanism is disposed on the load-bearing beam; a GPS positioning device is also disposed on the side surface of the load-bearing plate, and the GPS positioning device is used to obtain the real-time position of the main unit.

[0013] Preferably, both walking mechanisms include a first driving device, a driving wheel, and a driven wheel; the first driving device is fixedly installed at the end of the load-bearing plate away from the ground, and the output end of the first driving device is fixedly connected to the rotation shaft of the driving wheel; the driven wheel is rotatably connected to the main unit; both the driving wheel and the driven wheel are rollingly connected to a track on one side of the main unit.

[0014] Preferably, the hoisting mechanism includes a second drive unit, a hydraulic brake, a reducer, a drum, at least two water pipes, and a latch hook; the second drive unit is fixedly connected to the load-bearing beam, and a hydraulic brake is provided on the output end of the second drive unit; the drum is located on the edge of the load-bearing beam and is rotatably connected to the load-bearing beam; the reducer is straddling the output end of the second drive unit and the rotating shaft of the drum, and the input end of the reducer is connected to the output end of the second drive unit, and the output end of the reducer is connected to the rotating shaft of the drum; at least two water pipes are wound around the drum opposite to each other, one end of the at least two water pipes is connected to a water tank, and the other end of the at least two water pipes is wrapped around and fixed to one end of the latch hook, and the other end of the latch hook is hooked on the cleaning trolley unit; the at least two water pipes are made of flexible material and are used to supply water to the cleaning trolley unit; the hydraulic brake is used to lock the output end of the second drive unit; the second drive unit drives the drum through the reducer to drive the retraction and extension of the at least two water pipes, thereby adjusting the distance between the cleaning trolley unit and the main unit.

[0015] Preferably, the hydraulic brake includes a piston cylinder, a first piston rod, a second piston rod, an upper brake arm, a lower brake arm, an upper connecting rod, a brake stop block, a first pull arm, a second pull arm, and two brake pads. The piston cylinder is fixedly disposed relative to the second drive device, and the first piston rod and the second piston rod are symmetrically disposed at both ends of the piston cylinder. The movable ends on both sides of the piston cylinder are respectively hinged to one end of the first piston rod and one end of the second piston rod. The first pull arm and the second pull arm are disposed opposite to each other and spaced apart on both sides of the output shaft of the second drive device. The other end of the first piston rod is fixedly connected to one end of the first pull arm, and the upper connecting rod respectively hinges the ends of the first pull arm and the second pull arm near the first piston rod. The other end of the second pull arm is hinged to the second drive device. Brake pads are correspondingly disposed on the adjacent end faces of the first pull arm and the second pull arm. The brake pads are hinged to the first pull arm or the second pull arm, and the shape of the brake pads matches the output shaft of the second drive device. The other end of the lower brake arm is rotatably connected to the other end of the second drive device and the other end of the first pull arm.

[0016] Preferably, the cleaning trolley unit includes a support frame, a boom, two cover plates, a cleaning positioning device, several wheels, a rotating cleaning brush, and an infrared distance sensor. The support frame is hollow inside, with several wheels and a rotating cleaning brush located at one end of the support frame near the photovoltaic panel. The cleaning positioning device is located inside the support frame, and two cover plates are also located inside the support frame. The adjacent ends of the two cover plates are angled together, and the non-adjacent ends of the two cover plates extend towards the edge of the support frame. The vertical distance between the adjacent ends of the two cover plates and the photovoltaic panel is greater than the vertical distance between the non-adjacent ends of the two cover plates and the photovoltaic panel. A boom spans the top of the support frame, and both ends of the boom are fixedly connected to the support frame. A latch hook is hinged to the boom at different positions. Water droplets supplied by at least two water pipes fall onto the surfaces of the two cover plates and splash onto the surface of the photovoltaic panel. The cleaning positioning device is used to position and plan the movement path of the support frame along the photovoltaic panel. Several wheels drive the support frame to move relative to the photovoltaic panel, and the rotating cleaning brush is used on the surface of the photovoltaic panel that the support frame passes over. The infrared distance sensor is located on the side of the support frame near the photovoltaic panel and is used to detect grooves or protrusions in the direction of movement of the support frame.

[0017] More preferably, several of the walking wheels are Mecanum wheels, and each walking wheel's rotation axis is correspondingly provided with a third drive device; a fourth drive device is provided on the rotation axis of the rotating cleaning brush, and both the third and fourth drive devices are fixedly mounted on the support frame.

[0018] On the other hand, the present invention also provides a method for using a distributed photovoltaic robot, comprising the following steps:

[0019] S1: Construct the above-mentioned distributed photovoltaic robot; equip the cleaning vehicle unit with a cleanliness detection device; equip the water tank with a pressure sensor to detect the water pressure in the water tank in real time;

[0020] S2: A grid-based environmental model is performed on the working environment of the distributed photovoltaic robot to obtain a grid map. Each grid cell in the grid map is assigned as either an obstacle zone or a free zone. A Cartesian coordinate system is established, with the unit length of the coordinate axis being the size of the grid cell. The grid position is represented by Cartesian coordinates (x, y). The grid numbering starts from the bottom left corner of the grid map and is added sequentially from bottom to top and from left to right. The relationship between the grid number and the grid position coordinates is: m = (x-1)*N+y, y = mod(m, N), x = int(m, N)+1; m is the current grid number; x and y are the position coordinates of the distributed photovoltaic robot in the Cartesian coordinate system; N represents the grid cell. Figure 1 The size of the dimensional space is N0×N0; mod() is the division modulo operation; int() is the floor function. Let p1 be the initial point, then the expression for the optimal path between the initial point and the target point is as follows:

[0021] Where minf() is the extremum objective function; p i For non-initial path points on the optimal path, j = 2, 3, 4, ..., n; d(p i p i-1 ) is the Euclidean distance between two adjacent path points; P is the set of path points on the optimal path; W is the set of all grid points; O is the set of grid points that constitute obstacles, and the optimal path does not pass through obstacles;

[0022] S3: A particle swarm optimization algorithm improved by Cauchy inverse learning algorithm is used to train the algorithm and output the global optimal solution of the optimal path between the initial point and the target point.

[0023] S4: Substitute the global optimal solution into the expression for the optimal path between the initial point and the target point. The distributed photovoltaic robot plans the real-time positions of the host unit and the cleaning vehicle unit according to the expression for the optimal path, so as to realize the actions of opening and closing the water pipe and cleaning the photovoltaic panels.

[0024] S5: The cleaning trolley unit runs along the optimal path on the photovoltaic panel for 2-3 cycles. The cleanliness of the photovoltaic panel is detected by the cleanliness detection device. The cleaning ends when the cleanliness meets the standard.

[0025] Preferably, the particle swarm optimization algorithm improved by Cauchy inverse learning algorithm described in step S3, which trains the algorithm and outputs the global optimal solution of the optimal path between the initial point and the target point, specifically includes the following:

[0026] S31: Read the raster map data and make it have N0×N0 nodes, with the number of particles in each node ranging from [0, 1];

[0027] S32: Set the parameters for the particle swarm optimization algorithm;

[0028] S33: Initialize the particle swarm; let point M = x(x1, x2, x3, ..., x...). d ) lies in d-dimensional space, where x I ∈[a I b I ], a I and b I Let be the minimum and maximum values ​​of point M in the I-th dimension, respectively, where I = 1, 2, 3, ..., d; and let OM = x′(x′1, x′2, x′3, ..., x′) be the inverse point of point M. d ), where x′ I =a I +b I +x I If I = 1, 2, 3, ..., d; then the Cauchy reversal point at point M is... The Cauchy antiparticle group of the initial particle group is constructed using the Cauchy inversion point, and the Cauchy antiparticle group and the initial particle group are used as the final initial particle group; initial random positions and velocities are assigned to all particles in the final initial group.

[0029] S34: Update the particle's velocity and position; the velocity update equation is as follows: The equation for updating the particle's position is as follows: Where the subscript j represents the j-th particle, j = 1, 2, 3, ..., m; and Let represent the velocities of the j-th particle in the k-th and (k+1)-th generations in the d-th dimension, respectively. and These represent the positions of the j-th particle in the k-th and (k+1)-th generations in the d-th dimension, respectively. This represents the optimal position of the j-th particle in the d-th dimension during the k-th generation; c1 represents the optimal position of the particle swarm in the d-th dimension of the k-th generation; c2 is the inertial weight of the particle tracking its own historical individual optimal value; c3 is the inertial weight of the particle tracking the optimal value of the entire swarm; ω is the inertial weight of the j-th particle in the d-th dimension of the k-th generation; ζ and η are random constants uniformly distributed in the interval [0, 1]; γ is the constraint factor, with a value of 1.

[0030] S35: Decode the particle to obtain the particle's position vector p; sort the particle's position vector p to obtain the sorting J; obtain the path between the starting point and the ending point through the sorting J, and calculate the path length.

[0031] S36: If the iteration is complete, output the result; if the iteration is not complete, return to step S33.

[0032] More preferably, in step S34, the inertial weight c1 is set to 1.8, the inertial weight c2 is set to 1.7, the population size is 50, and the maximum number of iterations is 150.

[0033] The distributed photovoltaic robot and its usage method provided by this invention have the following advantages compared to the prior art:

[0034] (1) This application consists of a main unit, a cleaning trolley unit, a water tank and a hoisting mechanism. The cleaning trolley unit adopts an independent power source, which drives the Mecanum wheel to move the rotating cleaning brush. The Mecanum wheel can achieve axial and tangential sliding. Compared with the traditional cleaning method, it can achieve high-efficiency and fast cleaning of photovoltaic panels.

[0035] (2) In terms of route planning, when the cleaning vehicle unit is working, the infrared distance sensor can detect whether there is a groove or a protrusion in the direction of movement so that the cleaning vehicle unit can avoid the position. A heuristic search algorithm is adopted to evaluate the position of the 8 nodes adjacent to the current position and replan the movement route of the cleaning vehicle unit. Global positioning is achieved by GPS positioning device to realize real-time position tracking, thus realizing the high applicability of the distributed photovoltaic robot.

[0036] (3) Compared with traditional cleaning methods, this application can reduce working at heights and potential dangers, and reduce the safety risks to workers, which is especially important for solar power plants in remote or harsh environments; the distributed photovoltaic robot can also operate under temperature conditions while exposed to pollutants, avoiding the health threats to operators when using manual operation. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is a perspective view of the distributed photovoltaic robot and its usage method of the present invention;

[0039] Figure 2 This is a perspective view of a portion of the walking mechanism of the distributed photovoltaic robot and its usage method of the present invention.

[0040] Figure 3 This is a perspective view of the hydraulic brake of the distributed photovoltaic robot and its usage method of the present invention.

[0041] Figure 4This is a perspective view of the cleaning vehicle unit of the distributed photovoltaic robot and its usage method of the present invention;

[0042] Figure 5 This is a flowchart illustrating the process of the distributed photovoltaic robot and its usage method according to the present invention.

[0043] Figure 6 This is a flowchart of the particle swarm optimization algorithm improved by the Cauchy inverse learning algorithm for the distributed photovoltaic robot and its usage method of the present invention.

[0044] Reference numerals: 100, track; 200, main unit; 300, water tank; 400, cleaning trolley unit; 500, hoisting mechanism; 201, load-bearing plate; 202, load-bearing beam; 203, GPS positioning device; 204, traveling mechanism; 205, first drive device; 206, drive wheel; 207, driven wheel; 501, second drive device; 502, hydraulic brake; 503, reducer; 504, drum; 505. Water pipe; 506, latch hook; 5021, piston cylinder; 5022, first piston rod; 5023, second piston rod; 5024, upper brake arm; 5025, lower brake arm; 5026, upper connecting rod; 5027, brake stop block; 5028, first pull arm; 5029, second pull arm; 401, support frame; 402, lifting rod; 403, cover plate; 404, cleaning positioning device; 405, traveling wheel; 406, rotating cleaning brush. Detailed Implementation

[0045] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0046] like Figure 1 As shown, in one aspect, the present invention provides a distributed photovoltaic robot, comprising:

[0047] The two tracks 100 are parallel to each other and spaced apart;

[0048] The main unit 200 is straddling the two tracks 100, and the main unit 200 moves linearly relative to the extending direction of the two tracks 100.

[0049] Water tank 300 is installed on main unit 200 to provide water for cleaning photovoltaic panels. In order to monitor the remaining water in water tank 300, a pressure sensor can be installed in water tank 300 to calculate the water level by pressure. When the water level meets the usage requirements, photovoltaic panel cleaning can be carried out normally. When the water level is low, the water outlet valve of the water tank is closed, and the main unit 200 moves along the track 100 to the water tank replenishment point to replenish water. After replenishment, it returns to the position where cleaning is not completed to continue cleaning.

[0050] The cleaning trolley unit 400 is disposed on the surface of the photovoltaic panel on one side of the main unit 200. The cleaning trolley unit 400 is used to clean the photovoltaic panel near the main unit 200. The cleaning is carried out by a combination of spraying and roller brushing, which has a good cleaning effect on stains and dust.

[0051] A hoisting mechanism 500 is located at the end of the main unit 200 away from the track 100 and is connected to the cleaning trolley unit 400. It guides water from the water tank 300 to the cleaning trolley unit 400 and changes the distance between the cleaning trolley unit 400 and the main unit 200. During the cleaning process, the cleaning trolley unit 400 is powered independently, and the hoisting mechanism 500 constrains the maximum reachable position of the cleaning trolley unit 400. After the cleaning trolley unit 400 completes cleaning along a preset path, the hoisting mechanism 500 retracts and resets the cleaning trolley unit 400.

[0052] See attached document Figure 5 The workflow of the distributed photovoltaic robot in this application consists of the following steps:

[0053] (1) When the distributed photovoltaic robot is in standby mode, the pressure sensor under the water tank will automatically detect the water tank pressure and calculate the water level. When the water level is higher than the preset value, the photovoltaic panel cleaning robot will start cleaning.

[0054] (2) The distributed photovoltaic robot obtains the cleaning status of the photovoltaic panels, locks the target cleaning position, and the trolley is powered by an independent battery. It moves the host unit 200 and the cleaning trolley unit 400 to the starting cleaning position through the built track.

[0055] (3) The hoisting mechanism 500 and the cleaning trolley unit 400 operate synchronously. At the same time, the water valve of the water tank 300 is opened, and water is sprayed onto the photovoltaic panel. After the cleaning trolley unit 400 runs 2-3 cleaning cycles on the photovoltaic panel, the cleaning degree is detected by the cleaning trolley unit 400. When the cleaning degree meets the standard, the cleaning ends.

[0056] (4) After cleaning is completed, the cleaning trolley unit 400 moves to the next target position and repeats the above process;

[0057] (5) When the pressure sensor under the water tank 300 detects that the water level is too low, the cleaning trolley unit 400 automatically stops cleaning and records the position that has not been cleaned. The main unit 200 drives the cleaning trolley unit 400 back to the water tank replenishment point to replenish water. After replenishing water, it returns to the position that has not been cleaned and continues cleaning.

[0058] (6) When the cleaning trolley unit 400 is running, it will also detect whether there are grooves or protrusions in the direction of movement. Grooves or protrusions indicate obstacles in areas other than photovoltaic panels. The cleaning trolley unit 400 will plan a path to avoid obstacles until all photovoltaic panels are completely cleaned.

[0059] like Figure 1 As shown, in order to move stably along the track 100 and obtain the relative position of the main unit 200, the main unit 200 includes a load-bearing plate 201, a load-bearing beam 202, and a GPS positioning device 203; the load-bearing plate 201 is disposed between the two tracks 100 and extends outward in a direction away from the two tracks 100, and the water tank 300 is disposed at the end of the load-bearing plate 201 away from the ground; two traveling mechanisms 204 are symmetrically disposed on both sides of the load-bearing plate 201 along the axial extension direction of the track 100, the two traveling mechanisms 204 are fixedly connected to the load-bearing plate 201, and the movable ends of the two traveling mechanisms 204 are tumblingly connected to the track 100; the load-bearing beam 202 is disposed at the end of the load-bearing plate 201 away from the two tracks 100, and the hoisting mechanism 500 is disposed on the load-bearing beam 202; the side surface of the load-bearing plate 201 is also provided with a GPS positioning device 203, which is used to obtain the real-time position of the main unit 200. The main unit 200 carries the water tank 300 on one hand and drives the cleaning trolley unit 400 to the set starting cleaning position on the other hand. During the movement, the position change is obtained through the GPS positioning device 203.

[0060] like Figure 1 Combination Figure 2 As shown, both walking mechanisms 204 include a first drive device 205, a drive wheel 206, and a driven wheel 207. The first drive device 205 is fixedly mounted on the end of the load-bearing plate 201 away from the ground, and its output end is fixedly connected to the rotation shaft of the drive wheel 206. The driven wheel 207 is rotatably connected to the main unit 200. Both the drive wheel 206 and the driven wheel 207 are rollingly connected to the track 100 on one side of the main unit 200. In this embodiment, the track 100 is made of I-beams, and the first drive device 205 is a geared motor.

[0061] like Figure 1As shown, the hoisting mechanism 500 includes a second drive unit 501, a hydraulic brake 502, a reducer 503, a drum 504, at least two water pipes 505, and a latch hook 506. The second drive unit 501 is fixedly connected to the load-bearing beam 202, and the hydraulic brake 502 is provided on the output end of the second drive unit 501. The drum 504 is located on the edge of the load-bearing beam 202 and is rotatably connected to the load-bearing beam 202. The reducer 503 is straddling the output end of the second drive unit 501 and the rotating shaft of the drum 504, and the input end of the reducer 503 is connected to the output end of the second drive unit 501, while the output end of the reducer 503 is connected to the rotating shaft of the drum 504. At least two water pipes 505 are wound around the roller 504, with one end of each pipe connected to the water tank 300 and the other end of each pipe wrapped around and fixed to one end of a latch hook 506, the other end of which is hooked onto the cleaning trolley unit 400. The water pipes 505 are made of flexible material and are used to supply water to the cleaning trolley unit 400. A hydraulic brake 502 is used to lock the output end of the second drive device 501. The second drive device 501 drives the roller 504 through a reducer 503 to move the water pipes 505 in and out, thereby adjusting the distance between the cleaning trolley unit 400 and the main unit 200. The second drive unit 501 is a motor. When the second drive unit 501 is working, the diameter of its output end differs significantly from the diameter of the roller 504. To reduce the roller's rotational speed, a reducer 503 is needed for proportional reduction, such as reducing the rotational speed of the second drive unit 501 from several thousand revolutions per minute to the roller 504's several revolutions per minute. To ensure the position holding effect after braking, a hydraulic brake 502 is used to lock the current position of the output end of the second drive unit 501. The water pipe 505 can both draw water from the water tank 300 and serve as an adjusting component for adjusting the length of the cleaning trolley unit 400. The water pipe itself has a certain degree of flexibility, better adapting to the need for the cleaning trolley unit 400 to move arbitrarily along the surface of the photovoltaic panel.

[0062] To ensure that the water pipe 505 is aligned with the roller 504, this embodiment also includes a wedge-shaped joint, a pressure roller, and a fixed pulley. The wedge-shaped joint has a U-shaped channel inside. The pressure roller and the fixed pulley form a pair of rollers. The gap between the U-shaped channel and the pair of rollers is used to limit the position of the water pipe 505, ensuring that the water pipe maintains a certain tension when it is extended and retracted, so that it will not drag on the surface of the photovoltaic panel or affect the normal movement of the cleaning trolley unit 400 on the surface of the photovoltaic panel.

[0063] Specifically, such as Figure 3As shown, the hydraulic brake 502 includes a piston cylinder 5021, a first piston rod 5022, a second piston rod 5023, an upper brake arm 5024, a lower brake arm 5025, an upper connecting rod 5026, a brake stop block 5027, a first pull arm 5028, a second pull arm 5029, and two brake pads. The piston cylinder 5021 is fixedly disposed relative to the second drive device, and the first piston rod 5022 and the second piston rod 5023 are symmetrically disposed at both ends of the piston cylinder 5021. The movable ends on both sides of the piston cylinder 5021 are respectively hinged to one end of the first piston rod 5022 and one end of the second piston rod 5023. The first pull arm 5028 and the second pull arm 5029 are disposed opposite to and spaced apart from the second drive device. The output shaft of the second drive device 501 is positioned on both sides; the other end of the first piston rod 5022 is fixedly connected to one end of the first pull arm 5028, and the upper connecting rod 5026 respectively hinges the ends of the first pull arm 5028 and the second pull arm 5029 near the first piston rod 5022, and the other end of the second pull arm 5029 is hinged to the second drive device 501; brake pads are correspondingly provided on the adjacent end faces of the first pull arm 5028 and the second pull arm 5029, and the brake pads are hinged to the first pull arm 5028 or the second pull arm 5029, and the shape of the brake pads matches the output shaft of the second drive device 501; the other end of the lower brake arm 5025 is rotatably connected to the other end of the second drive device 501 and the first pull arm 5028 respectively. During braking, the hydraulic cylinder piston actuates, causing the first piston rod 5022 and the second piston rod 5023 to move in the same direction. The first piston rod 5022 extends outward, rotating the upper brake arm 5024 and changing the distance between the first pull arm 5028 and the output shaft of the second drive device 501. As the first piston rod 5022 extends further, the upper brake arm 5024, the upper connecting rod 5026, and the first pull arm 5028 further deform, causing the second pull arm 5029 to rotate relative to the output shaft of the second drive device 501. This causes the two brake pads to press against the output shaft of the second drive device 501 sequentially, thus completing the hydraulic locking of the current output shaft of the second drive device 501. When the piston cylinder 5021 drives the first piston rod 5022 and the second piston rod 5023 to reset, the first pull arm 5028 and the second pull arm 5029 also reset accordingly.

[0064] like Figure 1 and Figure 4As shown, the cleaning trolley unit 400 includes a support frame 401, a hanging rod 402, two cover plates 403, a cleaning positioning device 404, several wheels 405, a rotating cleaning brush 406, and an infrared distance sensor. The support frame 401 is hollow inside. Several wheels 405 and a rotating cleaning brush 406 are provided at one end of the support frame 401 near the photovoltaic panel. The cleaning positioning device 404 is provided inside the support frame 401. Two cover plates 403 are also provided inside the support frame 401. The adjacent ends of the two cover plates 403 are set at an angle, and the non-adjacent ends of the two cover plates 403 extend towards the edge of the support frame 401. The vertical distance between the adjacent ends of the two cover plates 403 and the photovoltaic panel is greater than the vertical distance between the non-adjacent ends of the two cover plates 403 and the photovoltaic panel. The hanging rod 402 is strung across the top of the support frame 401, and the two ends of the hanging rod 402 are fixedly connected to the support frame 401. A latch hook 506 is hinged to a boom 402 at different positions; water droplets from at least two water pipes 505 fall onto the inclined surfaces of the two cover plates 403 and splash onto the surface of the photovoltaic panel; a cleaning positioning device 404 is used to position and plan the movement path of the support frame 401 along the photovoltaic panel; several wheels 405 drive the support frame 401 to move relative to the photovoltaic panel; a rotating cleaning brush 406 is used on the surface of the photovoltaic panel that the support frame 401 passes over; the rotating cleaning brush 406 has several flexible brush handles arranged on its surface for flexible contact with the surface of the photovoltaic panel to achieve a flexible cleaning effect; an infrared distance sensor is set on the side of the support frame 401 near the photovoltaic panel to detect grooves or protrusions in the direction of movement of the support frame 401. Grooves or protrusions are identified as obstacles in the photovoltaic panel area, and the path will be bypassed during subsequent path planning. The cleaning positioning device 404 is used to obtain the current trajectory position of the support frame 401.

[0065] In this embodiment, the walking wheels 405 are all Mecanum wheels, a type of omnidirectional moving wheel structure capable of translation and rotation in any direction. Each walking wheel 405 has a corresponding third drive device on its rotation axis; a fourth drive device is mounted on the rotation axis of the rotating cleaning brush 406. Both the third and fourth drive devices are fixedly mounted on the support frame 401. Both the third and fourth drive devices are motors. A dedicated power supply is also provided on the support frame 401 for use by the cleaning positioning device 404, the third drive device, and the fourth drive device.

[0066] Additionally, please refer to the appendix. Figure 6 The present invention also provides a method for using a distributed photovoltaic robot, comprising the following steps:

[0067] S1: Construct the above-mentioned distributed photovoltaic robot; equip the cleaning vehicle unit with a cleanliness detection device; equip the water tank with a pressure sensor to detect the water pressure in the water tank in real time.

[0068] S2: A grid-based environmental model is performed on the working environment of the distributed photovoltaic robot to obtain a grid map. Each grid cell in the grid map is assigned as either an obstacle zone or a free zone. A Cartesian coordinate system is established, with the unit length of the coordinate axis being the size of the grid cell. The grid position is represented by Cartesian coordinates (x, y). The grid numbering starts from the bottom left corner of the grid map and is added sequentially from bottom to top and from left to right. The relationship between the grid number and the grid position coordinates is: m = (x-1)*N+y, y = mod(m, N), x = int(m, N)+1; m is the current grid number; x and y are the position coordinates of the distributed photovoltaic robot in the Cartesian coordinate system; N represents the grid cell. Figure 1 The size of the dimensional space is N0×N0; mod() is the division modulo operation; int() is the floor function. Let p1 be the initial point, then the expression for the optimal path between the initial point and the target point is as follows:

[0069] Where minf() is the extremum objective function; p i For non-initial path points on the optimal path, i = 2, 3, 4, ..., n; d(p i p i-1 ) is the Euclidean distance between two adjacent path points; P is the set of path points on the optimal path; W is the set of all grid points; O is the set of grid points that constitute obstacles, and the optimal path does not pass through obstacles.

[0070] S3: The particle swarm optimization algorithm is improved by Cauchy's inverse learning algorithm. The algorithm is trained and outputs the global optimal solution of the optimal path between the initial point and the target point.

[0071] Specifically, this step includes the following:

[0072] S31: Read the raster map data and make it have N0×N0 nodes, with the number of particles in each node ranging from [0, 1];

[0073] S32: Set the parameters for the particle swarm optimization algorithm;

[0074] S33: Initialize the particle swarm; let point M = x(x1, x2, x3, ..., x...). d ) lies in d-dimensional space, where x I ∈[a I b I ], a I and b I Let be the minimum and maximum values ​​of point M in the I-th dimension, respectively, where I = 1, 2, 3, ..., d; and let OM = x′(x′1, x′2, x′3, ..., x′) be the inverse point of point M. d ), where x′ I =aI +b I +x I If I = 1, 2, 3, ..., d; then the Cauchy reversal point at point M is... The Cauchy antiparticle group of the initial particle group is constructed using the Cauchy inversion point, and the Cauchy antiparticle group and the initial particle group are used as the final initial particle group; initial random positions and velocities are assigned to all particles in the final initial group.

[0075] S34: Update the particle's velocity and position; the velocity update equation is as follows: The equation for updating the particle's position is as follows: Where the subscript j represents the j-th particle, j = 1, 2, 3, ..., m; and Let represent the velocities of the j-th particle in the k-th and (k+1)-th generations in the d-th dimension, respectively. and These represent the positions of the j-th particle in the k-th and (k+1)-th generations in the d-th dimension, respectively. This represents the optimal position of the j-th particle in the d-th dimension during the k-th generation; c1 represents the optimal position of the particle in the d-th dimension of the k-th generation; c2 is the inertial weight of the particle tracking its own historical individual optimal value, and c1 takes a value of 1.8; c2 is the inertial weight of the particle tracking the optimal value of the entire population, and c2 takes a value of 1.7; ω is the inertial weight of the j-th particle in the d-th dimension of the k-th generation; ζ and η are random constants uniformly distributed in the interval [0, 1]; γ is the constraint factor, and takes a value of 1; in this embodiment, the population size is set to 50, and the maximum number of iterations is 150.

[0076] S35: Decode the particle to obtain the particle's position vector p; sort the particle's position vector p to obtain the sorting J; obtain the path between the starting point and the ending point through the sorting J, and calculate the path length.

[0077] S36: If the iteration is complete, output the result; if the iteration is not complete, return to step S33.

[0078] S4: Substitute the global optimal solution into the expression of the optimal path between the initial point and the target point. The distributed photovoltaic robot plans the real-time position of the host unit 200 and the cleaning vehicle unit 400 according to the expression of the optimal path, so as to realize the actions of opening and closing the water pipe 505 and cleaning the photovoltaic panel.

[0079] S5: The cleaning trolley unit 400 runs along the optimal path on the photovoltaic panel for 2-3 cycles. The cleanliness of the photovoltaic panel is detected by the cleanliness detection device. When the cleanliness meets the standard, it means that the cleaning quality meets the requirements and the cleaning ends. If the cleanliness does not meet the requirements, step S5 is repeated until the set maximum cleaning time or the maximum number of cleaning retries is reached, and then the cleaning ends.

[0080] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A distributed photovoltaic robot, characterized in that, include: Two tracks (100) are parallel to each other and spaced apart; The main unit (200) is straddling the two tracks (100) and moves linearly relative to the extension direction of the two tracks (100); the main unit (200) includes a load-bearing plate (201), a load-bearing beam (202), a GPS positioning device (203) and two walking mechanisms (204). A water tank (300), installed on the main unit (200), is used to provide water for cleaning the photovoltaic panels; A cleaning trolley unit (400) is disposed on the surface of a photovoltaic panel on one side of the main unit (200), and the cleaning trolley unit (400) is used to clean the photovoltaic panel near the main unit (200); A hoisting mechanism (500) is located at one end of the main unit (200) away from the track (100) and connected to the cleaning trolley unit (400); it is used to guide water in the water tank (300) to the cleaning trolley unit (400) and change the distance between the cleaning trolley unit (400) and the main unit (200); The hoisting mechanism (500) includes a second drive unit (501), a hydraulic brake (502), a reducer (503), a drum (504), at least two water pipes (505), and a latch hook (506); the second drive unit (501) is fixedly connected to the load-bearing beam (202), and the hydraulic brake (502) is provided on the output end of the second drive unit (501); the drum (504) is located on the edge of the load-bearing beam (202) and is rotatably connected to the load-bearing beam (202); the reducer (503) is straddling the output end of the second drive unit (501) and the rotating shaft of the drum (504), and the input end of the reducer (503) is connected to the output end of the second drive unit (501), and the output end of the reducer (503) is connected to the rotating shaft of the drum (504). Shaft connection; at least two water pipes (505) are wound around the roller (504) opposite to each other, one end of the at least two water pipes (505) is connected to the water tank (300), and the other end of the at least two water pipes (505) is wrapped around and fixed to one end of the latch hook (506), and the other end of the latch hook (506) is hooked on the cleaning trolley unit (400); the at least two water pipes (505) are made of flexible material and are used to supply water to the cleaning trolley unit (400); the hydraulic brake (502) is used to lock the output end of the second drive device (501); the second drive device (501) drives the roller (504) through the reducer (503) to drive the at least two water pipes (505) to retract and extend, thereby adjusting the distance between the cleaning trolley unit (400) and the main unit (200); The method of using distributed photovoltaic robots includes the following steps: S1: Construct the above-mentioned distributed photovoltaic robot; equip the cleaning vehicle unit with a cleanliness detection device; equip the water tank with a pressure sensor to detect the water pressure in the water tank in real time; S2: Perform grid-based environmental modeling on the working environment of the distributed photovoltaic robot to obtain a grid map. Each grid cell in the grid map is assigned as an obstacle zone or a free zone. Establish a Cartesian coordinate system, with the unit length of the coordinate axes being the size of the grid cell. The grid cell position is determined by Cartesian coordinates (…). x, y This indicates that the grid numbers are added sequentially, starting from the bottom left grid of the grid map and proceeding from bottom to top and left to right. The relationship between the grid number and the grid position coordinates is as follows: m =( x -1)* N + y , y = mod (m, N ), x = int ( m , N +1; m It is the current grille number; x and y These are the position coordinates of the distributed photovoltaic robot in a Cartesian coordinate system. N Refers to the size of the one-dimensional space of a raster map ; mod ( ) represents the remainder operation in division; int ( ) represents the integer operation, let p Let 1 be the initial point, then the expression for the optimal path between the initial point and the target point is as follows: ,in The objective function is the extreme value objective function; These are non-initial path points on the optimal path. ; It is the Euclidean distance between two adjacent path points; It is the set of path points on the optimal path; It is the set of all grid points; It is the set of grid points that constitute obstacles, and the optimal path does not pass through obstacles; S3: A particle swarm optimization algorithm improved by Cauchy inverse learning algorithm is used to train the algorithm and output the global optimal solution of the optimal path between the initial point and the target point. S4: Substitute the global optimal solution into the expression of the optimal path between the initial point and the target point. The distributed photovoltaic robot plans the real-time position of the host unit (200) and the cleaning vehicle unit (400) according to the expression of the optimal path, so as to realize the actions of opening and closing the water pipe (505) and cleaning the photovoltaic panel. S5: The cleaning trolley unit (400) runs along the optimal path on the photovoltaic panel for 2-3 cycles. The cleanliness of the photovoltaic panel is detected by the cleanliness detection device. The cleaning ends when the cleanliness meets the standard.

2. The distributed photovoltaic robot according to claim 1, characterized in that, The load-bearing plate (201) is set between the two tracks (100) and extends outward in a direction away from the two tracks (100). The water tank (300) is set at the end of the load-bearing plate (201) away from the ground. Two walking mechanisms (204) are symmetrically set on both sides of the load-bearing plate (201) along the axial extension direction of the track (100). The two walking mechanisms (204) are fixedly connected to the load-bearing plate (201), and the movable ends of the two walking mechanisms (204) are tumblingly connected to the track (100). The load-bearing beam (202) is set at the end of the load-bearing plate (201) away from the two tracks (100), and the hoisting mechanism (500) is set on the load-bearing beam (202). A GPS positioning device (203) is also set on the side surface of the load-bearing plate (201). The GPS positioning device (203) is used to obtain the real-time position of the host unit (200).

3. The distributed photovoltaic robot according to claim 2, characterized in that, Both of the walking mechanisms (204) include a first driving device (205), a driving wheel (206), and a driven wheel (207); the first driving device (205) is fixedly installed at the end of the load-bearing plate (201) away from the ground, and the output end of the first driving device (205) is fixedly connected to the rotation shaft of the driving wheel (206); the driven wheel (207) is rotatably connected to the main unit (200); the driving wheel (206) and the driven wheel (207) on one side of the main unit (200) are both tumbledly connected to the track (100) on that side.

4. The distributed photovoltaic robot according to claim 1, characterized in that, The cleaning trolley unit (400) includes a support frame (401), a boom (402), two cover plates (403), a cleaning positioning device (404), several wheels (405), a rotating cleaning brush (406), and an infrared distance sensor. The support frame (401) is hollow inside. Several wheels (405) and a rotating cleaning brush (406) are provided at one end of the support frame (401) near the photovoltaic panel. The cleaning positioning device (404) is provided inside the support frame (401). Two cover plates (403) are also provided inside the support frame (401). The adjacent ends of the two cover plates (403) are set at an angle, and the non-adjacent ends of the two cover plates (403) extend towards the edge of the support frame (401). The vertical distance between the adjacent ends of the two cover plates (403) and the photovoltaic panel is greater than the vertical distance between the non-adjacent ends and the photovoltaic panel. Vertical distance; a hanger (402) spans the top of the support frame (401), and the two ends of the hanger (402) are fixedly connected to the support frame (401) respectively; a latch hook (506) is hinged to the hanger (402) at different positions; water droplets provided by at least two water pipes (505) fall on the surface of the two cover plates (403) and splash onto the surface of the photovoltaic panel; a cleaning positioning device (404) is used to position and plan the movement path of the support frame (401) along the photovoltaic panel; several wheels (405) drive the support frame (401) to move relative to the photovoltaic panel, and a rotating cleaning brush (406) is used to clean the surface of the photovoltaic panel that the support frame (401) passes over; an infrared distance sensor is set on the side of the support frame (401) close to the photovoltaic panel to detect the grooves or protrusions in the movement direction of the support frame (401).

5. The distributed photovoltaic robot according to claim 4, characterized in that, Several walking wheels (405) are Mecanum wheels, and each walking wheel (405) has a corresponding third drive device on its rotation shaft; a fourth drive device is provided on the rotation shaft of the rotating cleaning brush (406), and both the third drive device and the fourth drive device are fixedly mounted on the support frame (401).

6. The distributed photovoltaic robot according to claim 1, characterized in that, The particle swarm optimization algorithm improved by Cauchy inverse learning algorithm described in step S3 is used to train the algorithm and output the global optimal solution of the optimal path between the initial point and the target point. Specifically, it includes the following: S31: Read raster map data and make it have There are 12 nodes, and the number of particles in each node ranges from [0, 1]. S32: Set the parameters for the particle swarm optimization algorithm; S33: Initialize the particle swarm; set points lie in d 3D space, in which , and Points M In the I Minimum and maximum values ​​in dimensional space. ;point M The opposite point is ,in , Then point M The Cauchy reversal point at that location is The Cauchy antiparticle group of the initial particle group is constructed using the Cauchy inversion point, and the Cauchy antiparticle group and the initial particle group are used as the final initial particle group; initial random positions and velocities are assigned to all particles of the final initial group. S34: Update the particle's velocity and position; the velocity update equation is as follows: The equation for updating the particle's position is as follows: ; where subscript j Indicates the first j Particles ; and They represent the first j The first particle d The first dimension k generation and first k +1 generation speed; and They represent the first j The first particle d The first dimension k generation and first k +1 generation position; Indicates the first j The first particle d The first dimension k The optimal position of the generation; The particle swarm is represented by the first d The first dimension k The optimal position of the generation; It is the inertial weight of a particle tracking its own historical best value; It is the inertial weight of the particle that tracks the optimal value of the entire population; For the first j The first particle d The first dimension k The inertial weight of the generation; and It is a random constant that is uniformly distributed in the interval [0, 1]. This is a constraint factor, with a value of 1. S35: Decode the particle to obtain its position vector. p ; for the particle's position vector p Sort the data to obtain the sorted results. J By sorting J Obtain the path between the starting point and the ending point, and calculate the path length; S36: If the iteration is complete, output the result; if the iteration is not complete, return to step S33.

7. The distributed photovoltaic robot according to claim 6, characterized in that, The inertial weight mentioned in step S34 The value is 1.8, representing the inertia weight. The value is 1.7, the population size is 50, and the maximum number of iterations is 150.

Images