Path planning method for automatic hollow glass cleaning device

Abstract

The invention discloses a path planning method for an automatic hollow glass cleaning device. The path planning method comprises the following steps: a data acquisition and preprocessing step; a region division and coefficient calculation step; and a path generation and screening step. According to the method, the safety risk coefficient of each cleaning path is calculated, the electric quantity consumption condition during cleaning in each area is predicted according to the electric quantity consumption model of the cleaning device, and for the cleaning device working vertically or obliquely, if the electric quantity is lower than a safety threshold value after cleaning in a certain area and the area below the area is a dangerous area, the safety risk coefficient is increased; and finally, a path with the total cleaning difficulty coefficient and the safety risk coefficient being comprehensively optimal is selected, so that the safety in the cleaning process can be ensured while the cleaning effect is ensured.

Description

Technical Field

[0001] The invention relates to the field of hollow glass cleaning, and in particular to a path planning method for an automatic hollow glass cleaning device. Background Art

[0002] With the development of modern architecture, insulating glass is increasingly used in buildings. Insulating glass not only provides good heat and sound insulation, but also looks beautiful. However, due to the environment in which it is located, such as being exposed to the air, it is easily contaminated with dust and stains, and needs to be cleaned regularly. Manual cleaning of insulating glass has problems such as low efficiency and high risk (for high-rise buildings), so automatic cleaning devices came into being.

[0003] There are many technical solutions for path planning in the existing automatic cleaning devices for insulating glass. For example, a common existing technology is to use simple zone cleaning to clean the insulating glass in a fixed order (such as from left to right, from top to bottom), without considering the distribution of stains on the insulating glass and the difference in cleaning difficulty in different areas. Some other technologies take the existence of stains into consideration, but simply divide the glass into stained areas and stain-free areas, without in-depth analysis of factors such as the type of stains and the difficulty of cleaning. In addition, in terms of energy consumption, existing path planning rarely comprehensively considers the energy consumption differences of the cleaning device in different areas, such as the relationship between the cleaning area and the initial position of the cleaning device and the relationship between the area of ​​the area and the area of ​​the cleaning head. Summary of the invention

[0004] The object of the present invention is to provide a path planning method for an automatic hollow glass cleaning device to solve the problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A path planning method for an automatic hollow glass cleaning device comprises the following steps:

[0007] Step S1, data collection and preprocessing step: collecting insulating glass related data, cleaning device related data and insulating glass orientation data, processing the insulating glass image to obtain insulating glass related feature data and stain related data, wherein the stain related data includes distinguishing different types of stain areas and related information;

[0008] Step S2, area division and coefficient calculation step: divide the insulating glass into multiple areas, calculate the cleaning difficulty coefficient and energy consumption coefficient corresponding to each area, the cleaning difficulty coefficient is calculated based on stain-related factors and area shape-related factors, and the energy consumption coefficient is calculated based on the relationship between the area of ​​the area and the area of ​​the cleaning head and the relationship between the area and the initial position of the cleaning device;

[0009] Step S3, path generation and screening step: a search algorithm is used in combination with a heuristic function that considers the cleaning difficulty coefficient and the energy consumption coefficient to generate multiple cleaning paths, the total cleaning difficulty coefficient, the total energy consumption and the safety risk coefficient of each cleaning path are calculated, and feasible paths whose total energy consumption meets the power conditions of the cleaning device are screened out, and the path with the best total cleaning difficulty coefficient and safety risk coefficient is selected from the feasible paths as the path for the cleaning device to clean the insulating glass this time.

[0010] In the present invention, in step S1, the insulating glass related data includes the insulating glass image, and the cleaning device related data includes the cleaning head area data, the cleaning device power data and the cleaning device position data.

[0011] In the present invention, in step S1, the insulating glass image is processed including grayscale, Gaussian filtering, edge detection, and binarization operations to obtain the area and shape data of the insulating glass, and a multi-threshold segmentation method is used to obtain stain data, distinguish between stubborn stain areas and common stain areas, and obtain their respective area and position information.

[0012] In the present invention, in step S2, the calculation formula for the cleaning difficulty coefficient is:

[0013] Cd=α×S p +β×S t +γ×C s

[0014] Where S p is the area ratio of stubborn stains, α is the weight of stubborn stains, S t is the area ratio of common stains, β is the weight of common stains, C s is the regional shape complexity coefficient, and γ is the shape complexity coefficient weight.

[0015] In the present invention, in step S2, the energy consumption coefficient calculation formula is:

[0016] C e =μ×A r +ν×D r

[0017] Among them, A r is the ratio of the area to the cleaning head area, μ is the area-related weight, D r is the distance from the area to the initial position of the cleaning device, and ν is the distance weight.

[0018] In the present invention, in step S3, the search algorithm is a breadth-first search algorithm.

[0019] In the present invention, in step S3,

[0020] The total cleaning difficulty coefficient is calculated as follows:

[0021]

[0022] The total energy consumption is calculated as follows:

[0023]

[0024] in, is the cleaning difficulty coefficient of the i-th area in the path, is the cleaning difficulty weight of the ith region, is the energy consumption coefficient of the ith area in the path, E m is the energy consumption of the cleaning device moving along the path.

[0025] In the present invention, in step S3, the safety risk coefficient predicts the power consumption during cleaning in each area based on the power consumption model of the cleaning device. For a cleaning device that works vertically or inclined, if the power level in a certain area is lower than the safety threshold after cleaning and there is a dangerous area below the area, the safety risk coefficient is increased.

[0026] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes a path planning method for an automatic hollow glass cleaning device.

[0027] A computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the processor executes a path planning method for an automatic hollow glass cleaning device.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] 1. The present invention collects detailed data related to insulating glass, cleaning device and insulating glass orientation, and processes the insulating glass image to obtain insulating glass related feature data and stain related data, including distinguishing different types of stain areas and related information. In the area division and coefficient calculation step, the cleaning difficulty coefficient is calculated by considering stain related factors (such as the area proportion of different types of stains) and regional shape related factors, and then in the path generation and screening step, the optimal cleaning path is generated in combination with the cleaning difficulty coefficient, so that different types of stains and regional characteristics can be effectively cleaned and the cleaning effect can be improved.

[0030] 2. The present invention calculates the energy consumption coefficient corresponding to each area, and the energy consumption coefficient is calculated based on the relationship between the area of ​​the area and the area of ​​the cleaning head and the relationship between the area and the initial position of the cleaning device. In the path generation and screening steps, the energy consumption coefficient and the cleaning difficulty coefficient are taken as considerations of the heuristic function, multiple cleaning paths are generated, and feasible paths whose total energy consumption meets the power conditions of the cleaning device are screened out, thereby optimizing energy consumption and avoiding unnecessary energy consumption.

[0031] 3. The present invention calculates the safety risk coefficient of each cleaning path, and predicts the power consumption during cleaning in each area according to the power consumption model of the cleaning device. For vertically or inclined cleaning devices, if the power level in a certain area is lower than the safety threshold after cleaning and there is a dangerous area below the area, the safety risk coefficient is increased. Finally, the optimal path with the total cleaning difficulty coefficient and the safety risk coefficient is selected. This ensures the safety of the cleaning process while ensuring the cleaning effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 The present invention is a schematic flow chart of a path planning method for an automatic hollow glass cleaning device. DETAILED DESCRIPTION

[0033] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0034] See also Figure 1 , the present invention provides a technical solution:

[0035] A path planning method for an automatic hollow glass cleaning device comprises the following steps:

[0036] Step S1, data collection and preprocessing step: collecting insulating glass related data, cleaning device related data and insulating glass orientation data, processing the insulating glass image to obtain insulating glass related feature data and stain related data, wherein the stain related data includes distinguishing different types of stain areas and related information;

[0037] In this step,

[0038] Step S101, collecting insulating glass images, cleaning device data and insulating glass orientation data.

[0039] Step S102, analyze the insulating glass image. First, grayscale processing is performed to convert the color image into a grayscale image; Gaussian filtering is applied to reduce noise, and then edge detection is performed; the image after edge detection is binarized, and the contour is searched to obtain the area data and shape data of the insulating glass; for the stain data, a multi-threshold segmentation method is used. In addition to detecting the stain area, the stubborn stain area and the common stain area are distinguished according to the stain color depth, texture and other characteristics. The stubborn stain area and the common stain area are marked on the insulating glass image respectively, and their respective area and position information are obtained.

[0040] Step S2, area division and coefficient calculation step: divide the insulating glass into multiple areas, calculate the cleaning difficulty coefficient and energy consumption coefficient corresponding to each area, the cleaning difficulty coefficient is calculated based on stain-related factors and area shape-related factors, and the energy consumption coefficient is calculated based on the relationship between the area of ​​the area and the area of ​​the cleaning head and the relationship between the area and the initial position of the cleaning device;

[0041] In this step,

[0042] Step S201, dividing the insulating glass into m areas.

[0043] Step S202, calculate the cleaning difficulty coefficients corresponding to the m regions. Cleaning difficulty coefficient Cd = α × S p +β×S t +γ×C s , where S p is the area ratio of stubborn stains (stubborn stain area divided by the total area of ​​the region), α is the stubborn stain weight (set according to experience, for example, α = 0.6); S t is the proportion of common stain area (common stain area divided by the total area of ​​the region), β is the weight of common stain (e.g. β = 0.3); C s is the area shape complexity coefficient (determined by the number of sides and curvature of the area shape, for example, a circular area C s =1, the polygon area is calculated based on the number of sides and angles, γ is the shape complexity coefficient weight, for example γ = 0.1).

[0044] In a specific embodiment, suppose there is a piece of insulating glass with stubborn stains on some areas. For example, due to long-term exposure to polluted environment, there are stains formed by dried cement dust on the glass surface near the construction site. The area of ​​the stubborn stains is divided by the total area of ​​the insulating glass to obtain the stubborn stain area ratio S. p. Since stubborn stains are difficult to clean, they have a greater impact on the cleaning difficulty coefficient. Give it a weight α, which represents the relative importance of stubborn stains in the overall cleaning difficulty. For example, if α = 0.6, it means that the proportion of stubborn stain area accounts for a large proportion in calculating the cleaning difficulty coefficient. In addition to stubborn stains, there are also some common stains such as dust and water stains on the glass. For example, in daily use, some dust will be absorbed on the glass surface. Similarly, calculate the proportion of the area of ​​common stains to the total area of ​​the insulating glass to get S t . Ordinary stains are relatively easy to clean, so the weight β is small. Assuming β = 0.3, its contribution to the cleaning difficulty coefficient after multiplying it with the area ratio of ordinary stains is relatively small compared to stubborn stains. Insulating glass may have different shapes, such as rectangular, circular or curved. For areas with complex shapes, the cleaning head of the cleaning device may be difficult to fully cover and require more adjustment operations. By analyzing the shape of insulating glass, a regional shape complexity coefficient C is defined s , for example, for a curved area, C s The value of may be high. Then multiply it by the weight γ (assuming γ = 0.1), quantify the influence of shape complexity on cleaning difficulty and add it to the calculation of cleaning difficulty coefficient.

[0045] Step S203, calculate the energy consumption coefficients corresponding to the m regions. e =μ×A r +ν×D r , where A r is the ratio of the area of ​​the region to the area of ​​the cleaning head (area of ​​the region divided by the area of ​​the cleaning head), μ is the area-related weight (e.g. μ = 0.7); D r is the distance from the area to the initial position of the cleaning device, and ν is the distance weight (eg ν = 0.3).

[0046] In a specific embodiment, assuming that the cleaning head is a circle with a fixed diameter, the insulating glass has a large rectangular area to be cleaned. Calculate the ratio of the area to the cleaning head area A r If the area of ​​the area is much larger than the cleaning head area, it means that the cleaning device needs to move multiple times in this area to fully cover it, consuming more energy. We give this ratio a weight μ (assuming μ = 0.7) to indicate the importance of the relationship between the area of ​​the area and the cleaning head area to the energy consumption coefficient. In addition, the initial position of the cleaning device is in the lower left corner of the insulating glass, and there is a stain area in the upper right corner. The distance D from the measurement area to the initial position of the cleaning device r The cleaning device needs to move a longer distance to reach this area, which consumes more energy. Give this distance a weight ν (assuming ν = 0.3), quantify the impact of the distance factor on energy consumption and add it to the calculation of the energy consumption coefficient.

[0047] Step S3, path generation and screening step: using a search algorithm combined with a heuristic function that considers the cleaning difficulty coefficient and the energy consumption coefficient to generate multiple cleaning paths, calculating the total cleaning difficulty coefficient, total energy consumption and safety risk coefficient of each cleaning path, screening out feasible paths whose total energy consumption meets the power condition of the cleaning device, and selecting the path with the best total cleaning difficulty coefficient and safety risk coefficient from the feasible paths as the path for the cleaning device to clean the insulating glass this time;

[0048] In this step,

[0049] Step S301, a breadth-first search algorithm is used in combination with a heuristic function to generate n cleaning paths. The heuristic function considers the cleaning difficulty coefficient and the energy consumption coefficient, and preferentially selects an area with low cleaning difficulty and low energy consumption as the next exploration area.

[0050] In a specific embodiment, it is assumed that the insulating glass is divided into a plurality of small square areas, each of which has its cleaning difficulty coefficient and energy consumption coefficient. The cleaning device is located at a corner of the insulating glass. First, the initial position of the cleaning device is used as the starting node of the search. Then, the breadth-first search algorithm is used to expand the search layer by layer starting from the starting node. When expanding the nodes, a heuristic function is combined to evaluate each possible expansion direction. The heuristic function takes into account the cleaning difficulty coefficient and the energy consumption coefficient. For example, for two adjacent areas A and B, if the sum of the cleaning difficulty coefficient and the energy consumption coefficient of area A is less than that of area B, then area A is more likely to be explored first during the search process. By continuously expanding the nodes, multiple cleaning paths from the starting position to various areas of the insulating glass are generated.

[0051] Step S302, calculating the total cleaning difficulty coefficient C of each of the n paths d-total , total energy consumption C e-total and safety risk factor C s Total cleaning difficulty coefficient in, is the cleaning difficulty coefficient of the i-th area in the path, is the cleaning difficulty weight of the ith area (set according to the area order and importance). in, is the energy consumption coefficient of the ith area in the path, E m The energy consumption of the cleaning device during movement in the path (calculated based on the moving distance and the energy consumption coefficient of the cleaning device). Safety risk factor C sCalculation of: Based on the power consumption model of the cleaning device, predict the power consumption of each area during cleaning. For vertical or inclined cleaning devices, if the power is lower than the safety threshold after cleaning a certain area and there is a dangerous area below the area (such as the ground or other areas that are prone to damage), the safety risk factor is increased. For example, if the power is lower than the safety threshold and there is a dangerous area below, C s =C S +1, otherwise C s =C s .

[0052] In a specific embodiment, assume that three cleaning paths have been generated, which are marked as path 1, path 2 and path 3. For each path, the total cleaning difficulty coefficient is calculated. For example, in path 1, there are three areas, and the cleaning difficulty coefficient of each area is calculated separately. Multiply by the corresponding cleaning difficulty weight Then add them together to get the total cleaning difficulty coefficient of path 1. Calculate the total energy in is the energy consumption coefficient of the ith area in the path, E m is the energy consumption of the cleaning device moving in the path. For example, in path 2, calculate the energy consumption coefficient of each area, add the energy consumed by the cleaning device moving on this path, and get the total energy consumption of path 2. Calculate the safety risk factor. Assume that the power consumption model of the cleaning device is known, and calculate it according to the power consumption of the cleaning device when cleaning in each area. For example, for a vertical cleaning device, if the power is lower than the safety threshold after cleaning a certain area and the area below is a dangerous area (such as there are fragile items below), then increase the safety risk factor. Filter out the total energy consumption C e-total The path that is less than the power data of the cleaning device is taken as a feasible path. Among the feasible paths, select the total cleaning difficulty coefficient C d-total The smallest path with the lowest safety risk factor is used as the path for the cleaning device to clean the insulating glass this time.

[0053] Step S303: calculate the total energy consumption C in n paths e-total The path that is smaller than the power data of the cleaning device is marked as a feasible path.

[0054] Step S304: Select the total cleaning difficulty coefficient C in the feasible path. d-total Minimum and safe risk factor C S The lowest path is used as the path for the cleaning device to clean the insulating glass.

[0055] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes a path planning method for an automatic hollow glass cleaning device.

[0056] A computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the processor executes a path planning method for an automatic hollow glass cleaning device.

[0057] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0058] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A path planning method for an automatic hollow glass cleaning device, characterized in that: The following steps are involved: Step S1, data collection and preprocessing step: collecting insulating glass related data, cleaning device related data and insulating glass orientation data, processing the insulating glass image to obtain insulating glass related feature data and stain related data, wherein the stain related data includes distinguishing different types of stain areas and related information; Step S2, area division and coefficient calculation step: divide the insulating glass into multiple areas, calculate the cleaning difficulty coefficient and energy consumption coefficient corresponding to each area, the cleaning difficulty coefficient is calculated based on stain-related factors and area shape-related factors, and the energy consumption coefficient is calculated based on the relationship between the area of ​​the area and the area of ​​the cleaning head and the relationship between the area and the initial position of the cleaning device; Step S3, path generation and screening step: a search algorithm is used in combination with a heuristic function that considers the cleaning difficulty coefficient and the energy consumption coefficient to generate multiple cleaning paths, the total cleaning difficulty coefficient, the total energy consumption and the safety risk coefficient of each cleaning path are calculated, and feasible paths whose total energy consumption meets the power conditions of the cleaning device are screened out, and the path with the best total cleaning difficulty coefficient and safety risk coefficient is selected from the feasible paths as the path for the cleaning device to clean the insulating glass this time.

2. A path planning method for an automatic hollow glass cleaning device according to claim 1, characterized in that: In the step S1, the insulating glass related data includes the insulating glass image, and the cleaning device related data includes the cleaning head area data, the cleaning device power data and the cleaning device position data.

3. A path planning method for an automatic hollow glass cleaning device according to claim 1, characterized in that: In step S1, the insulating glass image is processed including grayscale, Gaussian filtering, edge detection, and binarization operations to obtain the area and shape data of the insulating glass, and a multi-threshold segmentation method is used to obtain stain data, distinguish between stubborn stain areas and common stain areas, and obtain their respective area and position information.

4. A path planning method for an automatic hollow glass cleaning device according to claim 1, characterized in that: In step S2, the calculation formula for the cleaning difficulty coefficient is: Cd=α×S p +β×S t +γ×C s Where S p is the area ratio of stubborn stains, α is the weight of stubborn stains, S t is the area ratio of common stains, β is the weight of common stains, C s is the regional shape complexity coefficient, and γ is the shape complexity coefficient weight.

5. The path planning method of an automatic hollow glass cleaning device according to claim 1, characterized in that: In step S2, the energy consumption coefficient calculation formula is: C e =μ×A r +ν×D r Among them, A r is the ratio of the area to the cleaning head area, μ is the area-related weight, D r is the distance from the area to the initial position of the cleaning device, and ν is the distance weight.

6. A path planning method for an automatic hollow glass cleaning device according to claim 1, characterized in that: In step S3, the search algorithm is a breadth-first search algorithm.

7. The path planning method of an automatic hollow glass cleaning device according to claim 1, characterized in that: In the step S3, The total cleaning difficulty coefficient is calculated as follows: The total energy consumption is calculated as follows: in, is the cleaning difficulty coefficient of the i-th area in the path, is the cleaning difficulty weight of the ith region, is the energy consumption coefficient of the ith area in the path, E m is the energy consumption of the cleaning device moving along the path.

8. The path planning method of an automatic hollow glass cleaning device according to claim 1, characterized in that: In step S3, the safety risk coefficient predicts the power consumption during cleaning in each area based on the power consumption model of the cleaning device. For a cleaning device that works vertically or inclined, if the power in a certain area is lower than the safety threshold after cleaning and there is a dangerous area below the area, the safety risk coefficient is increased.

9. A computer device, characterized in that: It comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes a path planning method for an automatic cleaning device for insulating glass as described in any one of claims 1-8.

10. A computer-readable storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a processor, the processor executes a path planning method for an automatic cleaning device for insulating glass as described in any one of claims 1 to 8.

Images