A sweeping robot control system and method

Abstract

The present invention provides a sweeping robot control system and method, belonging to the technical field of sweeping robots, including: an image acquisition module, used to acquire a spatial image in a room; an image recognition module, used to recognize the spatial image, and obtain the area to be cleaned and garbage distribution information in the room; an instruction planning module, used to plan the sweeping robot's cleaning path according to the area to be cleaned, match the garbage distribution information with the cleaning path, configure garbage sub-information to each grid in the cleaning path, and configure the cleaning force to the corresponding grid to obtain a planning instruction set; a control module, used to control the sweeping robot to perform a cleaning operation according to the planning instruction set. By acquiring the spatial image in the room, performing image analysis, and obtaining the planning instruction set to control the sweeping robot to perform a cleaning operation, while realizing the path planning of the sweeping robot, the corresponding cleaning force is configured, which effectively improves the cleaning effect of the sweeping robot.

Description

Technical Field

[0001] The present invention relates to the technical field of sweeping robots, and in particular to a sweeping robot control system and method. Background Art

[0002] A sweeping robot is an automated household appliance that can replace manual cleaning of the floor, providing convenience and improving efficiency. It generally uses brushing and vacuuming to absorb debris on the floor into its own garbage collection box, thus completing the function of floor cleaning.

[0003] At present, sweeping robots use their built-in passive response sensors such as laser and ultrasonic to detect the surrounding environment of the sweeping robot and plan the cleaning path of the sweeping robot. However, when the environment in the room is complex and there are many obstacles, the laser sensor or ultrasonic sensor cannot fully receive the laser signal or ultrasonic signal reflected back by the surrounding objects, which affects the path planning of the sweeping robot.

[0004] In addition, traditional sweeping robots use a unified mode to clean the room, which makes it difficult to clean areas with a high degree of dirt in the room, resulting in poor cleaning effects and low cleaning efficiency of the sweeping robots.

[0005] Therefore, a control system and method for a sweeping robot are proposed. Summary of the invention

[0006] In order to solve the above technical problems, the present invention provides a sweeping robot control system and method, which are used to solve the problem that the sweeping robot in traditional technology performs environmental detection based on passive response sensors, resulting in poor path planning, using a unified mode for cleaning, and having poor cleaning effect.

[0007] The present invention provides a sweeping robot control system, comprising:

[0008] An image acquisition module, used for acquiring spatial images in the room;

[0009] An image recognition module, used to recognize the spatial image and obtain information about the area to be cleaned and garbage distribution in the room;

[0010] An instruction planning module is used to plan a cleaning path of the cleaning robot according to the area to be cleaned, and match the garbage distribution information with the cleaning path to configure garbage sub-information for each grid in the cleaning path, and configure the cleaning intensity for the corresponding grid, so as to obtain a planning instruction set for the area to be cleaned;

[0011] The control module is used to control the sweeping robot to perform cleaning operations according to the planned instruction set.

[0012] Preferably, in the sweeping robot control system, the image acquisition module comprises:

[0013] An image acquisition unit, used to acquire multiple local images of the room by performing multi-angle image acquisition through a camera assembly arranged on the top of the room;

[0014] An image processing unit, used for extracting characteristic parameters of each local image to splice the local images and generate a spliced ​​local image;

[0015] The image fusion unit is used to perform fusion processing on the stitching area of ​​the stitching local images to generate the spatial image.

[0016] Preferably, in the sweeping robot control system, the image processing unit comprises:

[0017] An image processing subunit, used for performing noise reduction and deformity correction processing on each local image;

[0018] The feature extraction subunit is used to obtain the feature parameters of each processed image by:

[0019] Perform convolution operation on the variable scale Gaussian function and the local image function to obtain the scale space of the corresponding processed image;

[0020] According to the scale space, a first pixel point at a non-edge position in the corresponding processed image and a second pixel point adjacent to the first pixel point are selected, and the first pixel point and the second pixel point are spatially mapped and compared to lock feature points existing in all the first pixel points;

[0021] The vector sum is obtained according to the gradient information of each feature point, and the feature vector of the corresponding feature point is obtained;

[0022] Inputting the feature vector into a vector parameter construction model to obtain feature parameters of the corresponding processed image;

[0023] A feature matching subunit, used for performing feature matching on feature parameters of the processed image and marking the overlapped area of ​​the processed image;

[0024] The image stitching subunit is used to stitch multiple processed images according to the overlapping areas to generate a stitched local image.

[0025] Preferably, in the sweeping robot control system, the image fusion unit comprises:

[0026] The energy acquisition subunit is used to acquire the energy spectrum of the spliced ​​area in the spliced ​​local image and obtain the energy value of each pixel in the spliced ​​area;

[0027] The seam search subunit is used to lock the minimum pixel point and the second minimum pixel point in the splicing area, and based on each minimum pixel point, sequentially search for the first minimum pixel point in the adjacent pixels until the search of the splicing area is completed, and at the same time, based on each second small pixel point, sequentially search for the second minimum pixel point in the second adjacent pixels until the search of the splicing area is completed;

[0028] The seam construction subunit is used to mark all the first minimum pixel points and the second minimum pixel points in the splicing area to obtain a first seam based on each minimum pixel point and a second seam based on each second minimum pixel point;

[0029] A sum calculation subunit is used to calculate the sum of the energy values ​​of the pixels on each slit, select a third slit whose sum of energy values ​​is less than a preset sum, and perform a fusion process on the third slit to obtain an optimal slit;

[0030] The fusion processing subunit is used to perform fusion processing on the spliced ​​areas of the spliced ​​local images according to the optimal cutting seam to generate a spatial image.

[0031] Preferably, the sweeping robot control system, the image recognition module, comprises:

[0032] A graph analysis unit, used for setting coordinates of the spatial image and performing gray value analysis and shape contour analysis on pixels in the spatial image;

[0033] The image marking unit is used to determine the area to be cleaned and the garbage distribution information in combination with the fixed positions of the household appliances in the room, and mark the outline of the area to be cleaned and the distribution position under the garbage distribution information on the spatial image, and record it as an electronic plane map.

[0034] Preferably, the command planning module of the sweeping robot control system includes:

[0035] A cleaning range analysis unit, configured to obtain cleaning width information of the cleaning robot according to the cleaning coverage range of the cleaning robot and a preset cleaning overlap;

[0036] A rasterization processing unit, configured to perform rasterization processing on the area to be cleaned in the electronic plane map according to the cleaning width information of the cleaning robot, so as to obtain an electronic raster map;

[0037] The path planning unit is used to obtain the current grid position of the sweeping robot in the electronic grid map and generate the cleaning path; specifically, by:

[0038] Each grid in the electronic grid map is regarded as a neuron, and a corresponding activation value A is assigned to each neuron;

[0039]

[0040] Among them, 1 / q is the self-excitation value of the neuron, a e is the external excitation value of the neuron, and c is a constant; when the cleaning robot has cleaned the room area corresponding to the grid, the activation value of the neuron corresponding to the grid is set to 0;

[0041] Among them, when , the next moving direction of the sweeping robot is defined, where δ1 represents the excitation threshold, and And 0<δ1≤1, and n0 represents the number of grids;

[0042] According to the current neuron O i1 , calculate the activation values ​​of the adjacent s neighborhood neurons, and select the neuron corresponding to the maximum activation value as the neuron O at the next moment i1+1 ;

[0043]

[0044] Among them, A i1+1 For the neuron O i1+1 The corresponding activation value, For the neuron O i1+1 The self-motivation value, a e(i1+1) For the neuron O i1+1 The external incentive value; For the neuron O i1+1 The activation variance of the s adjacent neighboring neurons involved; [] indicates the integer symbol; Represented as the neuron O i1+1 until the neurons corresponding to each grid in the electronic grid map are traversed, and the grid movement trajectory corresponding to the neurons is recorded as the cleaning path.

[0045] Preferably, the sweeping robot control system, the instruction planning module further includes:

[0046] a difficulty determination unit, configured to determine the garbage sub-information of each grid in the cleaning path, and analyze the garbage cleaning difficulty of the garbage sub-information based on the grid cleaning direction and the second cleaning difficulty based on the deviation direction of the grid cleaning direction;

[0047] A cleaning force calculation unit, configured to calculate the cleaning force of the corresponding grid according to the garbage cleaning difficulty and the second cleaning difficulty and in combination with a current garbage storage influence coefficient of the corresponding cleaning robot;

[0048]

[0049] Among them, A1 is the garbage cleaning difficulty of the corresponding grid; Q2 j1 is the second cleaning difficulty corresponding to the j1-th deviation direction of the grid, m01 represents the number of deviation directions of the grid; (Q2 j1 ) max For all Q2 in the corresponding grid j1 The maximum cleaning difficulty in ; δ2 is the current garbage storage influence coefficient of the corresponding sweeping robot; Q1 is the cleaning intensity of the corresponding grid;

[0050] A sequence construction unit, configured to construct a cleaning sequence based on the cleaning path based on the cleaning force;

[0051] A force optimization unit, used to optimize the cleaning force according to the conversion loss coefficient of the cleaning robot under any two cleaning forces;

[0052] If min(|Q1 j2 -Q1 j2+1 |,|Q1 j2 -Q1 j2-1 |)>ε1, then for Q1 j2-1 、Q1 j2 and Q1 j2+1 Sorting, ε1 represents the intensity threshold; Q1 j2 is the cleaning strength of the j2th grid; Q1 j2+1 is the cleaning strength of the j2+1th grid; Q1 j2-1 is the cleaning intensity of the j2-1th grid; min is the minimum value symbol; if Q1 j2 is the minimum, then according to Q1 j2 Replace and get the optimized result, where

[0053] r j2 =max(Z j2,j2+1 ,Z j2-1,j2 )×min(|Q1 j2 -Q1 j2+1 |,|Q1 j2 -Q1 j2-1 |,|Q1 j2-1 -Q1 j2+1 |)

[0054] Among them, Z j2,j2+1 is the conversion loss coefficient between the j2th and j2+1th cleaning forces; Z j2-1,j2 is the conversion loss coefficient between the j2-1th and j2th cleaning forces; r j2 is the adjustment coefficient based on the j2-th cleaning force; max is the maximum value symbol;

[0055] Otherwise, from Q1 j2-1 、Q1 j2 and Q1 j2+1 Get the maximum value for Q1 j2-1 、Q1 j2 and Q1 j2+1 Perform maximum value replacement respectively to obtain the optimization result;

[0056] If min(|Q1 j2 -Q1 j2+1 |,|Q1 j2 -Q1 j2-1 |)≤ε1, the cleaning force of the corresponding grid remains unchanged;

[0057] The optimization results of each grid are converted into sub-instructions, and then a planning instruction set is constructed.

[0058] Preferably, the sweeping robot control system further includes a dynamic obstacle avoidance module, including:

[0059] Establishing a set unit, for establishing a solution space set V of the angular velocity and linear velocity of the cleaning robot when there are obstacles on the cleaning path of the cleaning robot;

[0060]

[0061] Wherein, v is the linear velocity variable of the sweeping robot; ω is the angular velocity variable of the sweeping robot; md(v, ω) is the minimum distance between the sweeping robot and obstacles in the cleaning path; v amax is the maximum linear acceleration of the sweeping robot; ω amax is the maximum angular acceleration of the sweeping robot; an interval acquisition unit, used to acquire the control parameter adjustment interval V of the dynamic obstacle avoidance algorithm at time t according to the solution space set V of the sweeping robot t ;

[0062] V t ={(v,ω)|v∈[v r -v amax t,v r +v amax t]∩ω∈[ω r -ω amax t,ω r +ω amax t]}

[0063] Among them, the actual linear velocity of the sweeping robot at time t is v r and the actual angular velocity is ω r ;

[0064] A discretization unit is used to adjust the interval V by controlling the control parameters of the dynamic obstacle avoidance algorithm. t Discretization processing is performed;

[0065] A selection unit is used to calculate the control parameter adjustment interval V by traversing the optimization function t Among all the parameters in, select the one with the highest score F(v, ω) as the optimal solution Best(v b ,ω b ); the optimization function F(v, ω) is as follows:

[0066] F(v,ω)=w1·θ(v,ω)+w2·d(v,ω)+w3·mcv(v,ω)+w4·dt(v,ω)

[0067] Among them, θ(v, ω) is the evaluation parameter of the movement direction angle of the sweeping robot, that is, the approximation between the angle information when the sweeping robot moves according to the control parameter (v, ω) and the original angle information when the sweeping robot moves according to the cleaning path; d(v, ω) is the evaluation parameter of the distance between the sweeping robot and the obstacle; mcv(v, ω) is the evaluation parameter of the execution speed of the sweeping robot for the control parameter (v, ω); dt(v, ω) is the evaluation parameter of the distance between the sweeping robot and the original point of the cleaning path when the sweeping robot moves according to the control parameter (v, ω); w1, w2, w3, and w4 are the weight values ​​corresponding to the evaluation parameter of the movement direction angle of the sweeping robot, the evaluation parameter of the distance between the sweeping robot and the obstacle, the evaluation parameter of the execution speed of the sweeping robot, and the evaluation parameter of the distance between the sweeping robot and the original point of the cleaning path, respectively;

[0068] An optimization unit is used to optimize the optimal solution Best(v b ,ω b )Optimize the cleaning path of the sweeping robot.

[0069] The present invention provides a method for controlling a sweeping robot, comprising:

[0070] Step 1: Collect spatial images in the room;

[0071] Step 2: Identify the spatial image and obtain the area to be cleaned and garbage distribution information in the room;

[0072] Step 3: According to the area to be cleaned, a cleaning path of the sweeping robot is planned, and the garbage distribution information is matched with the cleaning path to configure garbage sub-information for each grid in the cleaning path, and the cleaning intensity is configured for the corresponding grid, so as to obtain a planning instruction set for the area to be cleaned;

[0073] Step 4: Control the sweeping robot to perform cleaning operations according to the planned instruction set.

[0074] Compared with the prior art, the beneficial effects of the present application are as follows: by acquiring the spatial image in the room and performing image analysis, a planning instruction set is obtained to control the sweeping robot to perform cleaning operations. While realizing path planning for the sweeping robot, the corresponding cleaning intensity is configured according to the area to be cleaned and the garbage distribution information, thereby effectively improving the cleaning effect of the sweeping robot.

[0075] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0076] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Figure 1 A schematic diagram of the structure of a sweeping robot control system provided by the present invention;

[0078] Figure 2 The present invention provides a flow chart of a method for controlling a sweeping robot. DETAILED DESCRIPTION

[0079] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0080] Embodiment 1:

[0081] The embodiment of the present invention provides a sweeping robot control system, referring to Figure 1 ,include:

[0082] An image acquisition module, used for acquiring spatial images in the room;

[0083] Image recognition module, used to recognize spatial images and obtain information about the area to be cleaned and the distribution of garbage in the room;

[0084] The instruction planning module is used to plan the cleaning path of the sweeping robot according to the area to be cleaned, and match the garbage distribution information with the cleaning path to configure garbage sub-information for each grid in the cleaning path, and configure the cleaning intensity for the corresponding grid to obtain a planning instruction set for the area to be cleaned;

[0085] The control module is used to control the sweeping robot to perform cleaning operations according to the planned instruction set.

[0086] The working principle of the above-mentioned sweeping robot control system is as follows: the image acquisition module collects the spatial image in the room, the image recognition module recognizes the spatial image, obtains the area to be cleaned and the garbage distribution information in the room, the instruction planning module plans the cleaning path of the sweeping robot according to the area to be cleaned, and matches the garbage distribution information with the cleaning path to configure garbage sub-information for each grid in the cleaning path, and configures the cleaning intensity for the corresponding grid, obtains the planned instruction set for the area to be cleaned, and the control module controls the sweeping robot to perform cleaning operations according to the planned instruction set.

[0087] In one embodiment, the image acquisition module may be a plurality of camera components, which acquire a plurality of local images of the area in the room and perform corresponding processing to obtain a spatial image.

[0088] In one embodiment, the image recognition module determines the area to be cleaned and the garbage distribution information based on the spatial image and the fixed positions of the furniture in the room.

[0089] In one embodiment, the instruction planning module plans the cleaning path of the sweeping robot according to the area to be cleaned, matches the cleaning path with the garbage distribution information, configures garbage sub-information for each grid in the cleaning path, configures corresponding cleaning intensity according to the garbage sub-information, and obtains a planning instruction set according to the cleaning path, the garbage sub-information of the grid and the corresponding cleaning intensity.

[0090] In one embodiment, the control module controls the sweeping robot to perform cleaning operations according to the planned instruction set. Specifically, each time the sweeping robot reaches a grid on the cleaning path, the sweeping robot performs cleaning operations according to the corresponding cleaning intensity.

[0091] The beneficial effect of the above technology is that by acquiring the spatial image in the room and performing image analysis, a planning instruction set is obtained to control the sweeping robot to perform cleaning operations. While realizing path planning for the sweeping robot, the corresponding cleaning intensity is configured according to the area to be cleaned and the garbage distribution information, thereby effectively improving the cleaning effect of the sweeping robot.

[0092] Embodiment 2:

[0093] The embodiment of the present invention provides a sweeping robot control system, an image acquisition module, including:

[0094] An image acquisition unit, used to acquire multiple local images of the room by performing multi-angle image acquisition through a camera assembly arranged on the top of the room;

[0095] An image processing unit, used for extracting characteristic parameters of each local image to splice the local images and generate a spliced ​​local image;

[0096] The image fusion unit is used to perform fusion processing on the stitching area of ​​the stitching local images to generate a spatial image.

[0097] In the above embodiments, the image acquisition unit acquires images from multiple angles in the room by means of a camera assembly disposed on the top of the room, thereby obtaining a plurality of local images in the room; the image processing unit extracts characteristic parameters of each local image to stitch the local images together to generate a stitched local image; and the image fusion unit performs fusion processing on the stitching area of ​​the stitched local images to generate a spatial image.

[0098] The beneficial effects of the above technology are: through the image acquisition unit, the acquisition of local images in the room is realized, and through the image processing unit, the characteristic parameters of the local images are extracted, the local images are spliced, and the acquisition of the spliced ​​local images is realized, and the image fusion unit fuses the spliced ​​areas of the spliced ​​local images to generate a spatial image, thereby realizing the acquisition of the spatial image in the room through the image acquisition unit, image processing unit, and image fusion unit in the above-mentioned image acquisition module.

[0099] Embodiment 3:

[0100] The embodiment of the present invention provides a sweeping robot control system, an image processing unit, including:

[0101] An image processing subunit, used for performing noise reduction and deformity correction processing on each local image;

[0102] The feature extraction subunit is used to obtain the feature parameters of each processed image by:

[0103] Perform convolution operation on the variable scale Gaussian function and the local image function to obtain the scale space of the corresponding processed image;

[0104] According to the scale space, a first pixel point at a non-edge position in the corresponding processed image and a second pixel point adjacent to the first pixel point are selected, and the first pixel point and the second pixel point are spatially mapped and compared to lock the feature points existing in all the first pixel points;

[0105] The vector sum is obtained according to the gradient information of each feature point, and the feature vector of the corresponding feature point is obtained;

[0106] Input the feature vector into the vector parameter construction model to obtain the feature parameters of the corresponding processed image;

[0107] A feature matching subunit, used for performing feature matching on feature parameters of the processed image and marking the overlapped area of ​​the processed image;

[0108] The image stitching subunit is used to stitch multiple processed images according to the overlapping areas to generate a stitched local image.

[0109] In the above embodiments, the image processing subunit performs noise reduction and deformity correction on each local image, and extracts feature parameters of the image processed by the image processing subunit through the feature extraction subunit; the feature matching subunit performs feature matching based on the feature parameters of the processed image, and marks the overlapping areas of the processed images; the image stitching subunit stitches multiple processed images based on the overlapping areas to generate a stitched local image.

[0110] In the above embodiments, the feature extraction subunit obtains the scale space of the corresponding processed image by convolving the variable-scale Gaussian function with the local image function; according to the scale space, the first pixel point at the non-edge of the corresponding processed image and the second pixel point adjacent to the first pixel point are selected, the first pixel point and the second pixel point are spatially mapped and compared, and the feature points existing in all the first pixels are locked; the vector sum is obtained according to the gradient information of each feature point to obtain the feature vector of the corresponding feature point; the feature vector is input into the vector parameter construction model to obtain the feature parameters of the corresponding processed image.

[0111] The beneficial effects of the above technology are: through the image processing subunit, noise reduction and deformity correction processing of the local image are realized, redundant information in the local image is removed, the image visual effect is improved, and the subsequent units are convenient for feature extraction of the local image; the feature extraction subunit extracts the feature parameters of the processed image, and the feature matching subunit performs feature matching according to the feature parameters of the processed image, and marks the overlapping areas; the image stitching subunit performs stitching processing according to the overlapping areas of multiple processed images, thereby realizing the acquisition of the stitched local image.

[0112] In the above embodiment, the feature matching subunit obtains the feature parameters K={k1, k2, ..., k m}, m is the number of feature vectors of the processed image p1(x,y), and the feature parameters N={n1,n2,…,n l}, l is the number of feature vectors of the processed image p2(x,y), k i is the i-th eigenvector of the processed image p1(x,y), n j is the jth eigenvector of the processed image p2(x,y);

[0113] Traverse and calculate the feature vector k in the feature parameter K i The Euclidean distance from each eigenvector in the feature parameter N is obtained to obtain the Euclidean distance from the eigenvector k in the feature parameter N. iThe feature vector with the smallest Euclidean distance is matched, where the value range of i is [1, m] and the value range of j is [1, l];

[0114] According to the matched feature vector k i and the eigenvector n j , marked in image p1(x,y) and image p2(x,y) respectively.

[0115] In the above embodiment, the characteristic parameters K={k1, k2, ..., k m} in which the feature vector k i and the characteristic parameters N = {n1,n2,…,n l}, and obtain the Euclidean distance between the feature parameter N and the feature vector k i The eigenvector n with the smallest Euclidean distance j , and the feature vector k i With the eigenvector n j Matching; according to the matched feature vector k i and the eigenvector n j , mark the overlapping areas in the local image p1(x,y) and the local image p2(x,y) respectively, and obtain the overlapping areas of the local images.

[0116] In the above embodiment, the image stitching subunit determines that the processed image p3(x, y) and the processed image p4(x, y) have an overlapping area according to the marking result, and performs a position transformation on the image p4(x, y) based on the image p3(x, y) to obtain the position transformed image p'4(x', y');

[0117]

[0118] Among them, the local image p4(x,y)=[x,y,1] T , the local image after position transformation p'4(x',y')=[x',y',1] T ,a 11 、a 12 、a 21 、a 22 is the scale rotation factor, a 13 is the lateral displacement factor, a 23 is the longitudinal displacement factor, a 31 is the lateral deformation, a 32 is the longitudinal deformation, a 33 is a constant 1, x represents the horizontal coordinate, and y represents the vertical coordinate.

[0119] In the above embodiment, the image stitching subunit transforms the position of the local image p4(x,y) based on the local image p3(x,y) when the local image p3(x,y) and the local image p4(x,y) have an overlapping area, and obtains the transformed local image p'4(x',y').

[0120] In the above embodiment, the position transformation of the local image p4(x, y) includes rigid transformation, similarity transformation, affine transformation and perspective transformation. Among them, the rigid transformation is to rotate and translate the local image p4(x, y); the details are as follows:

[0121]

[0122] Among them, the local image p4(x,y)=[x,y,1] T , the local image p' after rigid transformation 41 (x',y')=[x1',y1',1] T ,β is the rotation angle.

[0123] The similarity transformation is to rotate, translate and scale the local image p4(x,y); the details are as follows:

[0124]

[0125] Among them, the local image p' after similarity transformation 42 (x',y')=[x2',y2',1] T ,μ is the scale transformation factor.

[0126] Affine transformation is done by 11 、a 12 、a 21 、a 22 、a 13 、a 23 The 6 degrees of freedom factor setting realizes the transformation of the local image p4(x,y). The perspective transformation includes the above three forms of transformation. 11 、a 12 、a 21 、a 22 、a 13 、a 23 、a 31 、a 32 The setting of the factors realizes the transformation processing of the local image p4(x,y).

[0127] Embodiment 4:

[0128] The embodiment of the present invention provides a sweeping robot control system, an image fusion unit, including:

[0129] The energy acquisition subunit is used to acquire the energy spectrum of the spliced ​​area in the spliced ​​local image and obtain the energy value of each pixel in the spliced ​​area;

[0130] The seam search subunit is used to lock the minimum pixel point and the second minimum pixel point in the splicing area, and based on each minimum pixel point, sequentially search for the first minimum pixel point in the adjacent pixels until the search of the splicing area is completed, and at the same time, based on each second small pixel point, sequentially search for the second minimum pixel point in the second adjacent pixels until the search of the splicing area is completed;

[0131] The seam construction subunit is used to mark all the first minimum pixel points and the second minimum pixel points in the splicing area to obtain a first seam based on each minimum pixel point and a second seam based on each second minimum pixel point;

[0132] A sum calculation subunit is used to calculate the sum of the energy values ​​of the pixels on each slit, select a third slit whose sum of energy values ​​is less than a preset sum, and perform a fusion process on the third slit to obtain an optimal slit;

[0133] The fusion processing subunit is used to perform fusion processing on the spliced ​​areas of the spliced ​​local images according to the optimal cutting seam to generate a spatial image.

[0134] In the above embodiments, the minimum pixel point refers to a point with the smallest pixel value, and the second smallest pixel point refers to a point with a pixel value just higher than the minimum pixel value.

[0135] In the above embodiments, the slits can be effectively obtained by searching for points.

[0136] In the above embodiments, the fusion process refers to the operation of fusing slits with similar positions together to obtain one slit, and independently retaining slits with different positions, etc., to obtain the best slit.

[0137] The beneficial effects of the above technology are: through the energy acquisition subunit, the energy value of each pixel in the stitching area is acquired; through the slit search subunit, the slit construction subunit and the calculation subunit, the optimal slit is acquired, so that the difference of the pixels at the acquired optimal slit is small, and the fusion processing subunit fuses the stitching area according to the optimal slit, thereby acquiring the spatial image and effectively enhancing the quality and visual effect of the spatial image.

[0138] Embodiment 5:

[0139] The embodiment of the present invention provides a sweeping robot control system, an image recognition module, including:

[0140] An image analysis unit, used to set the coordinates of the spatial image and perform gray value analysis and shape contour analysis on the pixels in the spatial image;

[0141] The image marking unit is used to determine the area to be cleaned and the garbage distribution information in combination with the fixed positions of the household appliances in the room, and mark the outline of the area to be cleaned and the distribution position under the garbage distribution information on the spatial image, and record it as an electronic plane map.

[0142] In the above embodiments, the image analysis unit performs spatial image coordinate setting, pixel grayscale analysis and shape contour analysis, and the image marking unit marks the contour of the area to be cleaned and the distribution position of the garbage distribution information on the spatial image to construct an electronic floor plan map of the room area.

[0143] In the above embodiments, the gray value analysis refers to the gray distribution of pixels.

[0144] In the above embodiments, shape contour analysis refers to the contours of garbage, furniture, etc. present in the spatial image.

[0145] The beneficial effects of the above technology are: through the image analysis unit, the spatial image coordinate setting, pixel grayscale analysis and shape contour analysis are realized; the image marking unit marks the outline of the area to be cleaned and the distribution position under the garbage distribution information on the spatial image according to the fixed position of the furniture deployment in the room and the garbage distribution information, thereby realizing the image recognition module to obtain the electronic floor map of the room area.

[0146] Embodiment 6:

[0147] The embodiment of the present invention provides a sweeping robot control system, an instruction planning module, including:

[0148] A cleaning range analysis unit, used to obtain the cleaning width information of the cleaning robot according to the cleaning coverage range of the cleaning robot and a preset cleaning overlap;

[0149] A rasterization processing unit, used to perform rasterization processing on the area to be cleaned in the electronic plane map according to the cleaning width information of the sweeping robot, and obtain an electronic raster map;

[0150] The path planning unit is used to obtain the current grid position of the sweeping robot in the electronic grid map and generate a cleaning path; specifically, it is done by:

[0151] Each grid in the electronic grid map is regarded as a neuron, and each neuron is assigned a corresponding activation value A;

[0152]

[0153] Among them, 1 / q is the self-excitation value of the neuron, a e is the external excitation value of the neuron, and c is a constant; when the sweeping robot has cleaned the room area corresponding to the grid, the activation value of the neuron corresponding to the grid is set to 0;

[0154] Among them, when , the next moving direction of the sweeping robot is defined, where δ1 represents the excitation threshold, and And 0<δ1≤1, and n0 represents the number of grids;

[0155] According to the current neuron O i1 , calculate the activation values ​​of the adjacent s neighborhood neurons, and select the neuron corresponding to the maximum activation value as the neuron O at the next moment i1+1 ;

[0156]

[0157] Among them, A i1+1 For neuron O i1+1 The corresponding activation value, For neuron O i1+1 The self-motivation value, a e(i1+1) For neuron O i1+1 The external incentive value; For neuron O i1+1 The activation variance of the s adjacent neighboring neurons involved; [] indicates the integer symbol; Represented as neuron O i1+1 until the neurons corresponding to each grid in the electronic grid map are traversed, and the grid movement trajectory corresponding to the neuron is recorded as the cleaning path.

[0158] In the above embodiments, the cleaning range analysis unit obtains the cleaning width information of the sweeping robot according to the cleaning coverage range of the sweeping robot and the preset cleaning overlap; the rasterization processing unit rasterizes the area to be cleaned in the electronic plane map according to the cleaning width information of the sweeping robot to obtain the electronic grid map; the path planning unit generates the cleaning path information according to the current grid position of the sweeping robot in the electronic grid map.

[0159] In the above embodiment, each grid in the electronic grid map is regarded as a neuron, and a corresponding activation value is assigned to each neuron, and the activation value is determined by the neuron's own excitation value and the external excitation value; the neuron corresponding to the current grid position of the sweeping robot in the electronic grid map is regarded as the neuron at the current moment. i1 , calculate the activation values ​​of the adjacent neighborhood neurons, and select the neuron corresponding to the maximum activation value as the neuron O at the next moment i1+1, until the neurons corresponding to each grid in the electronic grid map are traversed, the grid movement trajectory corresponding to the neuron is recorded, and the cleaning path information is obtained.

[0160] In the above embodiment, by setting the excitation threshold δ1, when When the robot vacuum cleaner is in motion, the next moving direction of the robot vacuum cleaner is planned.

[0161] In the above embodiments, the maximum activation value of a neuron is obtained by summing up its own excitation value, the external excitation value and the correction factor and taking the maximum value.

[0162] The beneficial effects of the above technology are: through the cleaning range analysis unit, the cleaning width information of the sweeping robot is acquired; and the rasterization processing unit performs rasterization processing on the area to be cleaned in the electronic plane map to obtain the electronic grid map; the path planning unit generates the cleaning path information according to the current grid position of the sweeping robot in the electronic grid map, thereby realizing the function of system planning the cleaning path of the sweeping robot.

[0163] Embodiment 7:

[0164] The embodiment of the present invention provides a sweeping robot control system, an instruction planning module, and further includes:

[0165] A difficulty determination unit, used to determine the garbage sub-information of each grid in the cleaning path, and analyze the garbage sub-information based on the garbage cleaning difficulty of the grid cleaning direction and the second cleaning difficulty based on the deviation direction of the grid cleaning direction;

[0166] A cleaning force calculation unit, used to calculate the cleaning force of the corresponding grid according to the garbage cleaning difficulty and the second cleaning difficulty, and in combination with the current garbage storage influence coefficient of the corresponding sweeping robot;

[0167]

[0168] Among them, A1 is the garbage cleaning difficulty of the corresponding grid; Q2 j1 is the second cleaning difficulty corresponding to the j1-th deviation direction of the grid, m01 represents the number of deviation directions of the grid; (Q2 j1 ) max For all Q2 in the corresponding grid j1 The maximum cleaning difficulty in ; δ2 is the current garbage storage influence coefficient of the corresponding sweeping robot; Q1 is the cleaning intensity of the corresponding grid;

[0169] A sequence construction unit, used for constructing a cleaning sequence based on a cleaning path based on a cleaning force;

[0170] A force optimization unit, used to optimize the cleaning force of the sweeping robot according to the conversion loss coefficient under any two cleaning forces;

[0171] If min(|Q1 j2 -Q1 j2+1 |,|Q1 j2 -Q1 j2-1 |)>ε1, then for Q1 j2-1 、Q1 j2 and Q1 j2+1 Sorting, ε1 represents the intensity threshold; Q1 j2 is the cleaning strength of the j2th grid; Q1 j2+1 is the cleaning strength of the j2+1th grid; Q1 j2-1 is the cleaning intensity of the j2-1th grid; min is the minimum value symbol; if Q1 j2 is the minimum, then according to Q1 j2 Replace and get the optimized result, where

[0172] r j2 =max(Z j2，j2+1 , z j2-1，j2 )×min(|Q1 j2 -Q1 j2+1 |,|Q1 j2 -Q1 j2-1 |,|Q1 j2-1 -Q1 j2+1 |)

[0173] Among them, Z j2，j2+1 is the conversion loss coefficient between the j2th and j2+1th cleaning forces; Z j2-1，j2 is the conversion loss coefficient between the j2-1th and j2th cleaning forces; r j2 is the adjustment coefficient based on the j2-th cleaning force; max is the maximum value symbol;

[0174] Otherwise, from Q1 j2-1 、Q1 j2 and Q1 j2+1 Get the maximum value for Q1 j2-1 、Q1 j2 and Q1 j2+1 Perform maximum value replacement respectively to obtain the optimization result;

[0175] If min(|Q1 j2 -Q1 j2+1 |,|Q1 j2 -Q1 j2-1 |)≤ε1, the cleaning force of the corresponding grid remains unchanged;

[0176] The optimization results of each grid are converted into sub-instructions, and then a planning instruction set is constructed.

[0177] In the above embodiments, the garbage sub-information of each grid in the cleaning path is obtained through the difficulty determination unit, and the garbage sub-information is analyzed based on the garbage cleaning difficulty in the grid cleaning direction and the second cleaning difficulty in the deviation direction of the grid cleaning direction; the cleaning force calculation unit calculates the cleaning force of the corresponding grid according to the garbage cleaning difficulty and the second cleaning difficulty and the current garbage storage influence coefficient of the sweeping robot; the sequence construction unit constructs a cleaning sequence based on the cleaning path based on the cleaning force; the force optimization unit optimizes the cleaning force according to the conversion loss coefficient of the sweeping robot under any two cleaning forces, converts the optimization result of each grid into a sub-instruction, and constructs a planning instruction set.

[0178] In the above embodiments, the force optimization unit calculates the absolute value difference based on the cleaning force of three adjacent grids. When the difference is greater than the force threshold, the cleaning force of the three adjacent grids is sorted. When the cleaning force of the middle grid is the minimum value, the cleaning force of the middle grid is replaced and optimized according to the conversion loss coefficient and the adjustment coefficient. When the cleaning force of the remaining two grids is the minimum value, the maximum value is obtained from the cleaning force of the three adjacent grids, and the maximum value replacement optimization is performed respectively; when the difference is less than the force threshold, the cleaning force of the corresponding grid is kept unchanged.

[0179] In the above embodiments, garbage sub-information refers to the distribution of garbage in the corresponding grid, and analysis is performed by an image analysis model to analyze the grid image with garbage sub-information, so as to obtain the cleaning difficulty in different directions, and the model is pre-trained. Because, after path planning, the main cleaning direction of each grid has been determined, so, at this time, the cleaning difficulty in the existing deviation direction can be obtained through analysis, such as 1°, 3°, 10°, 30° and other directions away from the main cleaning direction. At this time, there will be m01 deviation directions and cleaning difficulties, that is, the more garbage there is, the denser and more difficult it is to clean up, the greater the corresponding cleaning difficulty, and the value range is (0,1).

[0180] In the above embodiment, the value range of the current garbage storage impact coefficient is (0, 1). The fuller the range, the larger the corresponding value.

[0181] In the above embodiment, the cleaning sequence = {cleaning intensity of grid 1, cleaning intensity of grid 2, ...}.

[0182] In the above embodiments, the conversion loss coefficient refers to the mode conversion loss from mode 1 to mode 2, the mode conversion loss from mode 1 to mode 3, and the like.

[0183] In the above embodiment, the planning instruction set = {sub-instructions of grid 1, sub-instructions of grid 2, ...}.

[0184] The beneficial effects of the above technology are: through the difficulty determination unit, the garbage cleaning difficulty based on the grid cleaning direction and the second cleaning difficulty based on the deviation direction of the grid cleaning direction for each grid in the cleaning path are obtained; the cleaning force calculation unit calculates the cleaning force of the corresponding grid according to the garbage cleaning difficulty, the second cleaning difficulty and the current garbage storage influence coefficient of the sweeping robot; through the sequence construction unit, based on the cleaning force, a cleaning sequence based on the cleaning path is constructed; the force optimization unit optimizes the cleaning force and constructs a planning instruction set, thereby realizing the function of the instruction planning module configuring the cleaning force for each grid in the cleaning path and obtaining the planning instruction set.

[0185] Since the garbage in the room is not evenly distributed, the traditional technology uses a unified light cleaning mode to clean the room, which makes the sweeping robot's cleaning effect on the room poor; the use of a unified deep cleaning module will make the cleaning time longer and the sweeping robot consumes a lot of energy. The above technology proposes matching the garbage sub-information of each grid in the cleaning path with the corresponding cleaning intensity, which can effectively solve the problem of using a single mode to clean the room in traditional technology, and effectively improve the room cleaning effect of the sweeping robot.

[0186] Embodiment 8:

[0187] The embodiment of the present invention provides a sweeping robot control system, a dynamic obstacle avoidance module, including:

[0188] Establishing a set unit for establishing a solution space set V of the angular velocity and linear velocity of the sweeping robot when there are obstacles on the sweeping path of the sweeping robot;

[0189]

[0190] Where v is the linear velocity variable of the sweeping robot; ω is the angular velocity variable of the sweeping robot; md(v, ω) is the minimum distance between the sweeping robot and the obstacles in the cleaning path; v amax is the maximum linear acceleration of the sweeping robot; ω amax is the maximum angular acceleration of the sweeping robot; an interval acquisition unit is used to obtain the control parameter adjustment interval V of the dynamic obstacle avoidance algorithm at time t according to the solution space set V of the sweeping robot t ;

[0191] V t ={(v,ω)|v∈[v r -v amax t,v r +v amax t]∩ω∈[ω r -ω amax t,ω r +ω amax t]}

[0192] Among them, the actual linear velocity of the sweeping robot at time t is v r and the actual angular velocity is ω r ;

[0193] Discretization unit, used to adjust the interval V by controlling the control parameters of the dynamic obstacle avoidance algorithm t Discretization processing is performed;

[0194] Select a unit to calculate the control parameter adjustment interval V through the optimization function traversal t Among all the parameters in, select the one with the highest score F(v, ω) as the optimal solution Best(v b ,ω b ); the optimization function F(v, ω) is as follows:

[0195] F(v,ω)=w1·θ(v,ω)+W2·d(v,ω)+W3·mcv(v,ω)+w4·dt(v,ω)

[0196] Among them, θ(v, ω) is the evaluation parameter of the moving direction angle of the sweeping robot, that is, the approximation between the angle information when the sweeping robot moves according to the control parameter (v, ω) and the original angle information when it moves according to the cleaning path; d(v, ω) is the evaluation parameter of the distance between the sweeping robot and the obstacle; mcv(v, ω) is the evaluation parameter of the execution speed of the sweeping robot for the control parameter (v, ω); dt(v, ω) is the evaluation parameter of the distance between the sweeping robot and the original point of the cleaning path when the sweeping robot moves according to the control parameter (v, ω); w1, w2, w3, and w4 are the weight values ​​corresponding to the evaluation parameter of the moving direction angle of the sweeping robot, the evaluation parameter of the distance between the sweeping robot and the obstacle, the evaluation parameter of the execution speed of the sweeping robot, and the evaluation parameter of the distance between the sweeping robot and the original point of the cleaning path, respectively;

[0197] Optimization unit, used to calculate the optimal solution Best(v b ,ω b )Optimize the cleaning path of the sweeping robot.

[0198] In the above embodiments, when the sweeping robot moves according to the cleaning path information, it detects that there are obstacles on the sweeping robot's cleaning path, and optimizes the sweeping robot's cleaning path through a dynamic obstacle avoidance algorithm to avoid obstacles and realize the system's dynamic obstacle avoidance function. The dynamic obstacle avoidance module is suitable for situations where the sweeping robot's working environment is in a dynamic change, and it takes a long time to replan the cleaning path. The dynamic obstacle avoidance module ensures the continuity of the sweeping robot's movement, and it can avoid obstacles in real time, optimize the cleaning path and perform corresponding movements. For example, when the sweeping robot detects that there is a user walking or standing on the cleaning path, it optimizes the cleaning path through a dynamic obstacle avoidance algorithm to avoid collision between the sweeping robot and the user.

[0199] In the above embodiment, based on the constraint that the sweeping robot does not collide with obstacles, a solution space set V of the angular velocity variable and the linear velocity variable of the sweeping robot is established by establishing a set unit; and the control parameter adjustment interval V is further solved by the interval acquisition unit based on the sweeping robot's own maximum linear acceleration and maximum angular acceleration limits. t The control parameter adjustment interval is discretized through the discretization unit, and the unit is selected to traverse and calculate all parameters in the control parameter adjustment interval through the optimization function to obtain the optimal solution Best(v b ,ω b ), the optimization unit is based on the optimal solution Best(v b ,ω b ) optimizes the cleaning path of the sweeping robot, and the control module controls the movement of the sweeping robot according to the cleaning path information optimized in the planning instruction set.

[0200] In this embodiment, the optimization function is calculated based on the evaluation parameters of the moving direction angle of the sweeping robot, the evaluation parameters of the distance between the sweeping robot and the obstacle, the evaluation parameters of the execution speed of the sweeping robot, the evaluation parameters of the distance between the sweeping robot and the original point of the cleaning path, and the corresponding weight values. The evaluation parameters of the distance between the sweeping robot and the original point of the cleaning path are used to evaluate the shortest distance between the sweeping robot and the original point of the cleaning path when the sweeping robot moves according to the control parameters (v, ω). The smaller the distance value, the higher the degree of fit between the sweeping robot and the original cleaning path.

[0201] The beneficial effects of the above technology are: the dynamic obstacle avoidance module is used to detect obstacles on the cleaning path of the sweeping robot, the cleaning path of the sweeping robot is optimized based on the dynamic obstacle avoidance algorithm, and the real-time avoidance of obstacles by the sweeping robot is realized, which solves the problem of the sweeping robot having to re-plan the cleaning path when encountering dynamic obstacles during the movement of the sweeping robot. While realizing real-time dynamic obstacle avoidance, the intelligence of the system controlling the sweeping robot is improved.

[0202] Embodiment 10:

[0203] The embodiment of the present invention provides a method for controlling a sweeping robot, which uses the sweeping robot control system described in Embodiment 1 to control the sweeping robot to perform a sweeping operation; Figure 2 ,include:

[0204] Step 1: Collect spatial images in the room;

[0205] Step 2: Identify the spatial image and obtain the area to be cleaned and the garbage distribution information in the room;

[0206] Step 3: According to the area to be cleaned, the cleaning path of the sweeping robot is planned, and the garbage distribution information is matched with the cleaning path to configure garbage sub-information for each grid in the cleaning path, and the cleaning intensity is configured for the corresponding grid to obtain a planning instruction set for the area to be cleaned;

[0207] Step 4: Control the sweeping robot to perform cleaning operations according to the planned instruction set.

[0208] The working principle of the above-mentioned sweeping robot control method is: collecting spatial images in the room, identifying the spatial images, obtaining the area to be cleaned and garbage distribution information in the room, planning the cleaning path of the sweeping robot according to the area to be cleaned, and matching the garbage distribution information with the cleaning path to configure garbage sub-information for each grid in the cleaning path, and configuring the cleaning intensity for the corresponding grid, obtaining the planned instruction set for the area to be cleaned, and the control module controls the sweeping robot to perform cleaning operations according to the planned instruction set.

[0209] The beneficial effect of the above technology is that by acquiring the spatial image in the room and performing image analysis, a planning instruction set is obtained to control the sweeping robot to perform cleaning operations. While realizing path planning for the sweeping robot, the corresponding cleaning intensity is configured according to the area to be cleaned and the garbage distribution information, thereby effectively improving the cleaning effect of the sweeping robot.

[0210] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A sweeping robot control system, characterized in that: include: An image acquisition module, used for acquiring spatial images in the room; An image recognition module is used to recognize the spatial image and obtain the area to be cleaned and garbage distribution information in the room; including: A graph analysis unit, used for setting coordinates of the spatial image and performing gray value analysis and shape contour analysis on pixels in the spatial image; A map marking unit, for determining the area to be cleaned and the garbage distribution information in combination with the fixed positions of the household appliances in the room, and marking the outline of the area to be cleaned and the distribution position under the garbage distribution information on the space image, and recording them as an electronic plane map; An instruction planning module is used to plan a cleaning path of the sweeping robot according to the area to be cleaned, and match the garbage distribution information with the cleaning path to configure garbage sub-information for each grid in the cleaning path, and configure the cleaning intensity for the corresponding grid, so as to obtain a planning instruction set for the area to be cleaned; including: A cleaning range analysis unit, configured to obtain cleaning width information of the cleaning robot according to the cleaning coverage range of the cleaning robot and a preset cleaning overlap; A rasterization processing unit, configured to perform rasterization processing on the area to be cleaned in the electronic plane map according to the cleaning width information of the cleaning robot, so as to obtain an electronic raster map; The path planning unit is used to obtain the current grid position of the sweeping robot in the electronic grid map and generate the cleaning path; specifically, by: Each grid in the electronic grid map is regarded as a neuron, and a corresponding activation value A is assigned to each neuron; Among them, 1 / q is the self-excitation value of the neuron, a e is the external excitation value of the neuron, and c is a constant; when the cleaning robot has cleaned the room area corresponding to the grid, the activation value of the neuron corresponding to the grid is set to 0; Among them, when , the next moving direction of the sweeping robot is defined, where δ1 represents the excitation threshold, and And 0<δ1≤1, and n0 represents the number of grids; According to the current neuron O i1 , calculate the activation values ​​of the adjacent s neighborhood neurons, and select the neuron corresponding to the maximum activation value as the neuron O at the next moment i1+1 ; Among them, A i1+1 For the neuron O i1+1 The corresponding activation value, For the neuron O i1+1 The self-motivation value, a e(i1+1) For the neuron O i1+1 The external incentive value; For the neuron O i1+1 The activation variance of the s adjacent neighboring neurons involved; [] indicates the integer symbol; Represented as the neuron O i1+1 until the neurons corresponding to each grid in the electronic grid map are traversed, and the grid movement trajectory corresponding to the neurons is recorded as the cleaning path; a difficulty determination unit, configured to determine the garbage sub-information of each grid in the cleaning path, and analyze the garbage cleaning difficulty of the garbage sub-information based on the grid cleaning direction and the second cleaning difficulty based on the deviation direction of the grid cleaning direction; A cleaning force calculation unit, configured to calculate the cleaning force of the corresponding grid according to the garbage cleaning difficulty and the second cleaning difficulty and in combination with a current garbage storage influence coefficient of the corresponding cleaning robot; Among them, A1 is the garbage cleaning difficulty of the corresponding grid; Q2 j1 is the second cleaning difficulty corresponding to the j1-th deviation direction of the grid, m01 represents the number of deviation directions of the grid; (Q2 j1 ) max For all Q2 in the corresponding grid j1 The maximum cleaning difficulty in ; δ2 is the current garbage storage influence coefficient of the corresponding sweeping robot; Q1 is the cleaning intensity of the corresponding grid; A sequence construction unit, configured to construct a cleaning sequence based on the cleaning path based on the cleaning force; A force optimization unit, used to optimize the cleaning force according to the conversion loss coefficient of the cleaning robot under any two cleaning forces; If min(|Q1 j2 -Q1 j2+1 |,|Q1 j2 -Q1 j2-1 |)>ε1, then for Q1 j2-1 、Q1 j2 and Q1 j2+1 Sorting, ε1 represents the intensity threshold; Q1 j2 is the cleaning strength of the j2th grid; Q1 j2+1 is the cleaning strength of the j2+1th grid; Q1 j2-1 is the cleaning intensity of the j2-1th grid; min is the minimum value symbol; if Q1 j2 is the minimum, then according to Q1 j2 Replace and get the optimized result, where r j2 =max(Z j2,j2+1 ,Z j2-1,j2 )×min(|Q1 j2 -Q1 j2+1 |,|Q1 j2 -Q1 j2-1 |,|Q1 j2-1 -Q1 j2+1 |) Among them, Z j2,j2+1 is the conversion loss coefficient between the j2th and j2+1th cleaning forces; Z j2-1,j2 is the conversion loss coefficient between the j2-1th and j2th cleaning forces; r j2 is the adjustment coefficient based on the j2-th cleaning force; max is the maximum value symbol; Otherwise, from Q1 j2-1 、Q1 j2 and Q1 j2+1 Get the maximum value for Q1 j2-1 、Q1 j2 and Q1 j2+1 Perform maximum value replacement respectively to obtain the optimization result; If min(|Q1 j2 -Q1 j2+1 |,|Q1 j2 -Q1 j2-1 |)≤ε1, the cleaning force of the corresponding grid remains unchanged; The optimization result of each grid is converted into sub-instructions, and then a planning instruction set is constructed; a control module is used to control the sweeping robot to perform a cleaning operation according to the planning instruction set.

2. The sweeping robot control system according to claim 1, characterized in that: The image acquisition module comprises: An image acquisition unit, used to acquire multiple local images of the room by performing multi-angle image acquisition through a camera assembly arranged on the top of the room; An image processing unit, used for extracting characteristic parameters of each local image to splice the local images and generate a spliced ​​local image; The image fusion unit is used to perform fusion processing on the stitching area of ​​the stitching local images to generate the spatial image.

3. The sweeping robot control system according to claim 2, characterized in that: The image processing unit comprises: An image processing subunit, used for performing noise reduction and deformity correction processing on each local image; The feature extraction subunit is used to obtain the feature parameters of each processed image by: Perform convolution operation on the variable scale Gaussian function and the local image function to obtain the scale space of the corresponding processed image; According to the scale space, a first pixel point at a non-edge position in the corresponding processed image and a second pixel point adjacent to the first pixel point are selected, and the first pixel point and the second pixel point are spatially mapped and compared to lock feature points existing in all the first pixel points; The vector sum is obtained according to the gradient information of each feature point, and the feature vector of the corresponding feature point is obtained; Inputting the feature vector into a vector parameter construction model to obtain feature parameters of the corresponding processed image; A feature matching subunit, used for performing feature matching on feature parameters of the processed image and marking the overlapped area of ​​the processed image; The image stitching subunit is used to stitch multiple processed images according to the overlapping areas to generate a stitched local image.

4. The sweeping robot control system according to claim 2, characterized in that: The image fusion unit comprises: The energy acquisition subunit is used to acquire the energy spectrum of the spliced ​​area in the spliced ​​local image and obtain the energy value of each pixel in the spliced ​​area; The seam search subunit is used to lock the minimum pixel point and the second minimum pixel point in the splicing area, and based on each minimum pixel point, sequentially search for the first minimum pixel point in the adjacent pixels until the search of the splicing area is completed, and at the same time, based on each second small pixel point, sequentially search for the second minimum pixel point in the second adjacent pixels until the search of the splicing area is completed; The seam construction subunit is used to mark all the first minimum pixel points and the second minimum pixel points in the splicing area to obtain a first seam based on each minimum pixel point and a second seam based on each second minimum pixel point; A sum calculation subunit is used to calculate the sum of the energy values ​​of the pixels on each slit, select a third slit whose sum of energy values ​​is less than a preset sum, and perform a fusion process on the third slit to obtain an optimal slit; The fusion processing subunit is used to perform fusion processing on the spliced ​​areas of the spliced ​​local images according to the optimal cutting seam to generate a spatial image.

5. The sweeping robot control system according to claim 1, characterized in that: Also included is a dynamic obstacle avoidance module, including: Establishing a set unit, for establishing a solution space set V of the angular velocity and linear velocity of the cleaning robot when there are obstacles on the cleaning path of the cleaning robot; Wherein, v is the linear velocity variable of the sweeping robot; ω is the angular velocity variable of the sweeping robot; md(v,ω) is the minimum distance between the sweeping robot and obstacles in the cleaning path; v amax is the maximum linear acceleration of the sweeping robot; ω amax is the maximum angular acceleration of the sweeping robot; an interval acquisition unit, used to acquire the control parameter adjustment interval V of the dynamic obstacle avoidance algorithm at time t according to the solution space set V of the sweeping robot t ; V t ={(v,ω)|v∈[v r -v amax t,v r +v amax t]∩ω∈[ω r -ω amax t,ω r +ω amax t]} Wherein, the actual linear velocity of the sweeping robot at time t is v r and the actual angular velocity is ω r ; A discretization unit is used to adjust the interval V by controlling the control parameters of the dynamic obstacle avoidance algorithm. t Discretization processing is performed; A selection unit is used to calculate the control parameter adjustment interval V by traversing the optimization function t Among all the parameters in, select the one with the highest score F(v,ω) as the optimal solution Best(v b ,ω b ); the optimization function F(v,ω) is as follows: F(v,ω)=w1·θ(v,ω)+w2·d(v,ω)+w3·mcv(v,ω)+w4·dt(v,ω)wherein, θ(v,ω) is the evaluation parameter of the movement direction angle of the sweeping robot, that is, the approximation between the angle information when the sweeping robot moves according to the control parameter (v,ω) and the original angle information when the sweeping robot moves according to the cleaning path; d(v,ω) is the evaluation parameter of the distance between the sweeping robot and the obstacle; mcv(v,ω) is the evaluation parameter of the execution speed of the sweeping robot for the control parameter (v,ω); dt(v,ω) is the evaluation parameter of the distance between the sweeping robot and the original point of the cleaning path when the sweeping robot moves according to the control parameter (v,ω); w1, w2, w3, and w4 are the weight values ​​corresponding to the evaluation parameter of the movement direction angle of the sweeping robot, the evaluation parameter of the distance between the sweeping robot and the obstacle, the evaluation parameter of the execution speed of the sweeping robot, and the evaluation parameter of the distance between the sweeping robot and the original point of the cleaning path, respectively; An optimization unit is used to optimize the optimal solution Best(v b ,ω b )Optimize the cleaning path of the sweeping robot.

6. A method for controlling a sweeping robot, characterized in that: The cleaning robot is controlled by the cleaning robot control system as claimed in claim 1 to perform cleaning operations; comprising: Step 1: Collect spatial images in the room; Step 2: Identify the spatial image and obtain the area to be cleaned and garbage distribution information in the room; Step 3: According to the area to be cleaned, a cleaning path of the sweeping robot is planned, and the garbage distribution information is matched with the cleaning path to configure garbage sub-information for each grid in the cleaning path, and the cleaning intensity is configured for the corresponding grid, so as to obtain a planning instruction set for the area to be cleaned; Step 4: Control the sweeping robot to perform cleaning operations according to the planned instruction set.

Images