Cleaning mode determination method and device, computer equipment and storage medium
By acquiring target scene data of cleaning robots and using pre-trained cleaning patterns to determine models, the problems of poor cleaning effect and energy waste of cleaning robots in different states and areas are solved, and more comprehensive cleaning pattern optimization is achieved.
Patent Information
- Application Number
- CN202410495327.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-04-23
- Publication Date
- 2025-10-24
AI Technical Summary
Existing cleaning robots suffer from poor cleaning performance and energy waste when faced with different self-states and areas to be cleaned.
By acquiring target scene data for the cleaning robot, including target environmental attribute data and machine attribute data of the area to be cleaned, and inputting it into a pre-trained cleaning mode determination model, the target cleaning mode is determined, and the cleaning parameters are optimized by taking into account the environment and machine status.
The cleaning effect is improved, energy waste is avoided, and a more comprehensive cleaning mode determination is achieved.
Smart Images
Figure CN120827313A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present disclosure relates to the technical field of cleaning robots, and particularly relates to a cleaning mode determination method and device, a computer device and a storage medium. BACKGROUND
[0002] A cleaning robot is a cleaning device that can clean hard floors, absorb and remove sewage. The use of cleaning robots in various fields of society has become very common, especially in places with wide hard floors such as stations, ports, airports, workshops, warehouses, schools, hospitals, hotels, and shopping malls. The concept of using machines instead of manpower for cleaning has been deeply rooted in people's minds. Recently, with the recognition of this new cleaning method by people, the demand for cleaning robots has surged.
[0003] In related technologies, the water discharge level of a cleaning robot is determined according to the environmental information of a cleaning area, and the determined water discharge level is used as a cleaning mode to clean the cleaning area. However, this cleaning method has the problems of poor cleaning effect and / or energy waste when the cleaning robot faces different machine states and different cleaning areas. SUMMARY
[0004] To overcome the problems in the related art, the present disclosure provides a cleaning mode determination method, device, computer device and storage medium.
[0005] According to a first aspect of embodiments of the present disclosure, the present disclosure provides a cleaning mode determination method, the method comprising:
[0006] obtaining target scene data of a cleaning robot, wherein the target scene data comprises target environmental attribute data of a cleaning area and target machine attribute data of the cleaning robot;
[0007] inputting the target scene data into a cleaning mode determination model to obtain a target cleaning mode output by the cleaning mode determination model; wherein the target cleaning mode comprises target values of a plurality of cleaning parameters in the cleaning robot.
[0008] According to any one of the embodiments of the present disclosure, the cleaning mode determination model is obtained by training in the following manner:
[0009] training a to-be-trained model to convergence according to a plurality of sample scene data in a training set to obtain the cleaning mode determination model, wherein the sample scene data comprises sample environmental attribute data of a sample cleaning area and sample machine attribute data of the cleaning robot, the sample scene data is labeled with a sample cleaning mode label, and the sample cleaning mode label comprises sample values of a plurality of cleaning parameters in the cleaning robot.
[0010] According to any one of the embodiments of the present disclosure, the sample scene data comprises a sample feature vector, and a label serial number of a sample cleaning mode label corresponding to the sample feature vector, and the sample feature vector is obtained by arranging the sample environment attribute data and the sample machine attribute data in a preset order.
[0011] According to any one of the embodiments of the present disclosure, the to-be-trained model comprises a decision tree model.
[0012] According to any one of the embodiments of the present disclosure, the training of the to-be-trained model to convergence according to the plurality of sample scene data in the training set to obtain the cleaning mode determination model comprises:
[0013] According to the plurality of sample scene data in the training set, a first information entropy is determined, and according to the sample environment attribute data and the sample machine attribute data of the plurality of sample scene data, a second information entropy corresponding to each attribute data is determined.
[0014] According to the first information entropy and the second information entropy corresponding to each attribute data, a classification priority of each attribute data is determined.
[0015] Based on the decision tree model, the classification priority of each attribute data, the plurality of sample scene data and the sample cleaning mode label are used to obtain the cleaning mode determination model.
[0016] According to any one of the embodiments of the present disclosure, the determination of the classification priority of each attribute data according to the first information entropy and the second information entropy corresponding to each attribute data comprises:
[0017] According to the first information entropy and the second information entropy corresponding to each attribute data, an information amount corresponding to each attribute data is determined.
[0018] All attribute data is arranged in descending order according to the corresponding information amount to obtain the classification priority of each attribute data.
[0019] According to any one of the embodiments of the present disclosure, the determination of the information amount corresponding to each attribute data according to the first information entropy and the second information entropy corresponding to each attribute data comprises:
[0020] The difference between the first information entropy and the second information entropy corresponding to each attribute data is used as the information amount corresponding to each attribute data.
[0021] According to any one of the embodiments of the present disclosure, the target environment attribute data comprises at least one of a weather state, an air state and a floor state of the to-be-cleaned area.
[0022] In combination with any of the embodiments of the present disclosure, the target machine attribute data includes at least one of power information, water information, and consumable loss information of the cleaning robot.
[0023] In a second aspect, the present disclosure also provides a cleaning mode determination apparatus, comprising:
[0024] a data acquisition module configured to acquire target scene data of the cleaning robot, wherein the target scene data includes target environment attribute data of a to-be-cleaned area and target machine attribute data of the cleaning robot;
[0025] a mode determination module configured to input the target scene data into a cleaning mode determination model to obtain a target cleaning mode output by the cleaning mode determination model; wherein the target cleaning mode includes target values of a plurality of cleaning parameters in the cleaning robot.
[0026] In one of the embodiments, the cleaning mode determination apparatus described above can further comprise:
[0027] a model training module configured to train the cleaning mode determination model in the following manner:
[0028] training a to-be-trained model to convergence according to a plurality of sample scene data in a training set to obtain the cleaning mode determination model, wherein the sample scene data includes sample environment attribute data of a sample to-be-cleaned area and sample machine attribute data of the cleaning robot, the sample scene data is labeled with a sample cleaning mode label, and the sample cleaning mode label includes sample values of a plurality of cleaning parameters in the cleaning robot.
[0029] In one of the embodiments, the sample scene data includes a sample feature vector and a label serial number of a sample cleaning mode label corresponding to the sample feature vector, and the sample feature vector is obtained by arranging the sample environment attribute data and the sample machine attribute data in a preset order.
[0030] In one of the embodiments, the to-be-trained model includes a decision tree model.
[0031] In one of the embodiments, the model training module described above can comprise:
[0032] an information entropy determination unit configured to determine a first information entropy according to a plurality of sample scene data in a training set, and determine a second information entropy corresponding to each attribute data according to sample environment attribute data and sample machine attribute data of the plurality of sample scene data;
[0033] a priority determination unit configured to determine a classification priority of each attribute data according to the first information entropy and the second information entropy corresponding to each attribute data.
[0034] The model determination unit is configured to determine the cleaning mode determination model based on a decision tree model according to the classification priority of each attribute data, the plurality of sample scene data, and the sample cleaning mode label.
[0035] In one of the embodiments, the priority determination unit can include:
[0036] The information amount determination subunit is configured to determine an information amount corresponding to each attribute data according to the first information entropy and a second information entropy corresponding to each attribute data.
[0037] The priority determination subunit is configured to arrange all attribute data in descending order according to the corresponding information amount to obtain the classification priority of each attribute data.
[0038] In one of the embodiments, the information amount determination subunit is specifically configured to:
[0039] The difference between the first information entropy and the second information entropy corresponding to each attribute data is taken as the information amount corresponding to each attribute data.
[0040] In one of the embodiments, the target environment attribute data includes at least one of a weather state, an air state, and a floor state of the area to be cleaned.
[0041] In one of the embodiments, the target machine attribute data includes at least one of power information, water information, and consumable consumption information of the cleaning robot.
[0042] In a third aspect, the present application provides a computer program product, including computer programs / instructions, which, when executed by a processor, implement the steps of the method of any of the embodiments.
[0043] In a fourth aspect, the present application provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the method of any of the embodiments when executing the program.
[0044] In a fifth aspect, the present application provides a computer-readable storage medium, which stores a computer program, and the program, when executed by a processor, implements the steps of the method of any of the embodiments.
[0045] The technical solutions provided by the embodiments of the present disclosure can include the following beneficial effects:
[0046] Compared with the related art, only the environmental factors of the to-be-cleaned area are considered, and the cleaning mode determined according to the environmental information of the to-be-cleaned area only includes the water output level of the cleaning robot. In the embodiment of the present disclosure, by considering the target environmental data of the to-be-cleaned area and the target machine attribute data of the cleaning robot, the target scene data of the cleaning robot is obtained, and then the target scene data is input into the pre-trained cleaning mode determination model, so that the target cleaning mode including the target values of multiple cleaning parameters of the cleaning robot can be more comprehensively and accurately determined, and the cleaning effect is improved when the cleaning robot faces different states and different to-be-cleaned areas, and the effect of avoiding energy waste is realized.
[0047] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present disclosure. BRIEF DESCRIPTION OF DRAWINGS
[0048] The accompanying drawings, which are incorporated into the specification and constitute a part of the present disclosure, illustrate embodiments consistent with the present disclosure and, together with the specification, serve to explain the principles of the present disclosure.
[0049] Figure 1 FIG. 1 is a flowchart of a cleaning mode determination method according to some example embodiments.
[0050] Figure 2 FIG. 2 is a flowchart of another cleaning mode determination method according to some example embodiments.
[0051] Figure 3 FIG. 3 is a flowchart of still another cleaning mode determination method according to some example embodiments.
[0052] Figure 4 FIG. 4 is a block diagram of a cleaning mode determination apparatus according to some example embodiments.
[0053] Figure 5 FIG. 5 is a hardware structure diagram of a computer device according to some example embodiments. DETAILED DESCRIPTION
[0054] The example embodiments will be described in detail herein, and examples thereof are shown in the accompanying drawings. When the following description refers to the accompanying drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following example embodiments do not represent all implementations consistent with the present disclosure. Instead, they are merely examples of apparatuses and methods consistent with some aspects of the present disclosure as detailed in the appended claims.
[0055] The terminology used in the present disclosure is for the purpose of describing particular embodiments only and is not intended to be limiting of the present disclosure. As used in the present disclosure and the appended claims, the singular forms "a," "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will also be understood that the term "and / or," as used herein, refers to and encompasses any and all possible combinations of one or more of the associated listed items.
[0056] It is to be understood that, although the terms first, second, third, etc. can be used herein to describe various information, the information should not be limited to these terms. These terms are only used to distinguish one piece of information from another. For example, a first information can also be termed a second information without departing from the scope of the present disclosure, similarly, a second information can also be termed a first information. Depending on the context, the word "if' as used herein can be interpreted as meaning "when" or "if upon a determination." For example, if a first condition is met, then a second condition can be met.
[0057] A cleaning robot is a cleaning device that cleans hard floors while sucking up dirty water and taking the dirty water away from the scene. The use of cleaning robots in various fields of society has become very common, especially in places with wide hard floors such as stations, ports, airports, workshops, warehouses, schools, hospitals, hotels, stores, etc. The concept of cleaning by machines instead of manpower has been deeply rooted in people's minds. Recently, with people's recognition of this new cleaning method of cleaning robots, the demand for cleaning robots has surged.
[0058] In the related art, the water discharge level of the cleaning robot is determined according to the environmental information of the area to be cleaned, and the determined water discharge level is used as a cleaning mode to clean the area to be cleaned. However, this cleaning method has the problem of poor comprehensiveness.
[0059] To solve the above problems, the present disclosure proposes a cleaning mode determination method, which is applied to a scenario of determining the cleaning mode of a cleaning robot in a case that the cleaning robot needs to clean different to-be-cleaned areas in different states. Optionally, the cleaning mode determination method can be executed by a cleaning mode determination system for determining the cleaning mode of the cleaning robot. Optionally, the cleaning mode determination system can be integrated locally on the cleaning robot; that is, the cleaning robot can acquire target scene data composed of target machine attribute data of the cleaning robot and target environment attribute data of the to-be-cleaned area through the cleaning mode determination system, and input the acquired target scene data into a pre-trained cleaning mode determination model to determine a target cleaning mode. Alternatively, the cleaning mode determination system can be integrated on a server, and can also run on a cloud or other network server; the cleaning mode determination system can interact with the cleaning robot through a network, acquire target scene data of the cleaning robot, determine a target cleaning mode according to the target scene data, and then send the target cleaning mode to the cleaning robot to control the cleaning robot to clean the to-be-cleaned area according to the target cleaning mode.
[0060] The first aspect of the present disclosure provides a cleaning mode determination method. Please refer to Figure 1 which comprises the following steps:
[0061] S101, acquiring target scene data of a cleaning robot.
[0062] The target scene data includes target environment attribute data of a to-be-cleaned area and target machine attribute data of the cleaning robot. The target environment attribute data can be environment attribute data obtained by the cleaning robot scanning the to-be-cleaned area; for example, the target environment attribute data can include at least one of weather state, air state and floor state of the to-be-cleaned area. The weather state can include illumination condition, weather condition, season, etc.; the air state can include air humidity, air quality, etc.; and the floor state can include floor material, floor dirtiness, etc. For example, the target environment attribute data can include good illumination, fine weather, spring, 30% air humidity, good air quality, marble ceramic tile as floor material, and relatively dirty floor dirtiness, etc. For example, the target machine attribute data includes at least one of power information (such as remaining power) of the cleaning robot, water information (such as remaining water) and consumable consumption information. For example, the target machine attribute data can include 56% remaining power, 40% remaining water, and 50% consumable consumption, etc. The to-be-cleaned area is an area that needs to be cleaned, and the to-be-cleaned area can be a local area (such as a specific area specified by a user) or the whole area of a map constructed by the cleaning robot.
[0063] Optionally, the target scene data composed of the target machine attribute data of the cleaning robot and the target environment attribute data of the to-be-cleaned area can be acquired through a locally set cleaning mode determination system. Alternatively, the target scene data composed of the target environment attribute data of the to-be-cleaned area and the target machine attribute data of the cleaning robot can be acquired by interacting with the cleaning robot through a network and receiving a cleaning mode determination request sent by the cleaning robot for the to-be-cleaned area.
[0064] S102, input the target scene data into the cleaning mode determination model to obtain a target cleaning mode output by the cleaning mode determination model.
[0065] The cleaning mode determination model can be a pre-trained neural network model for determining a cleaning mode. The target cleaning mode includes target values of a plurality of cleaning parameters of the cleaning robot. For example, the number of cleaning parameters included in the cleaning mode is limited, and the value corresponding to each cleaning parameter is also limited, so the number of cleaning modes composed of the values of the cleaning parameters is also limited. For example, the cleaning mode can include six parameters, i.e., the rotation speed of the side brush, the rotation speed of the roller, the motor suction, the water output, the voice volume, and the mechanical structure extension degree. Each cleaning parameter includes three values, i.e., low, medium, and high, which results in a limited number of cleaning modes. In the case of a limited number of cleaning modes, this step is used to select a cleaning mode from all cleaning modes as the target cleaning mode. For example, the target cleaning mode can be the cleaning mode shown in Table 1.
[0066] Table 1
[0067]
[0068] Optionally, after the target scene data of the cleaning robot is acquired, the target scene data can be input into a pre-trained cleaning mode determination model. The cleaning mode determination model outputs a target cleaning mode indicating the cleaning robot to clean the to-be-cleaned area, i.e., target values of a plurality of cleaning parameters.
[0069] Compared with the related art, only the environmental factors of the to-be-cleaned area are considered, and the cleaning mode determined according to the environmental information of the to-be-cleaned area only includes the water output level of the cleaning robot. In the embodiment of the present disclosure, by considering the target environmental data of the to-be-cleaned area and the target machine attribute data of the cleaning robot, the target scene data of the cleaning robot is obtained, and then the target scene data is input into the pre-trained cleaning mode determination model, so that the target cleaning mode including the target values of the multiple cleaning parameters of the cleaning robot can be more comprehensively and accurately determined, and the cleaning effect is improved when the cleaning robot faces different self-states and different to-be-cleaned areas, and the effect of avoiding energy waste is achieved.
[0070] On the basis of the above-mentioned embodiment, in one exemplary embodiment, the cleaning mode determination model can be trained in the following manner:
[0071] According to the multiple sample scene data in the training set, the to-be-trained model is trained to converge, and the cleaning mode determination model is obtained.
[0072] The training set can be a data set used to train the to-be-trained model to obtain the cleaning mode determination model. The sample scene data can be sample data in the training set used to train the to-be-trained model. The sample scene data includes sample environmental attribute data of a sample to-be-cleaned area and sample machine attribute data of the cleaning robot. The sample environmental attribute data can include at least one of weather state, air state and floor state of the sample to-be-cleaned area. The sample machine attribute data can include at least one of power information, water information and consumable loss information of the cleaning robot. The sample scene data is labeled with a sample cleaning mode label, and the sample cleaning mode label includes sample values of multiple cleaning parameters of the cleaning robot.
[0073] Optionally, the multiple sample scene data in the training set can be input into the to-be-trained model to train the to-be-trained model until the to-be-trained model converges, and the cleaning mode determination model is obtained. Specifically, the sample scene data can be input into the to-be-trained model to obtain predicted values corresponding to the sample scene data. Further, a loss value can be determined according to the predicted values corresponding to the sample scene data and the sample cleaning mode label corresponding to the sample scene data. The network parameters of the to-be-trained model are adjusted according to the loss value until the to-be-trained model meets the convergence condition (for example, the to-be-trained model is iterated for a preset number of times, and the loss value is less than a preset loss threshold), and the cleaning mode determination model is obtained.
[0074] It can be understood that by training the to-be-trained model until the to-be-trained model converges by using the plurality of sample scene data labeled with sample cleaning mode labels in the training set, the effect of efficiently obtaining a more accurate cleaning mode determination model can be achieved.
[0075] On the basis of the above-mentioned embodiments, in an exemplary embodiment, the sample scene data comprises a sample feature vector, and a label sequence number of a sample cleaning mode label corresponding to the sample feature vector. The sample feature vector is obtained by arranging the sample environment attribute data and the sample machine attribute data in a preset order.
[0076] The sample feature vector is obtained by arranging the plurality of attribute data in a preset arrangement order, and then writing the sample value corresponding to each attribute data in the arrangement order to generate the sample feature vector corresponding to the sample scene data. For example, in the case where the number of cleaning modes is limited, each cleaning mode can be taken as a sample cleaning mode label, and the sequence number of the cleaning mode can be taken as the corresponding label sequence number. The label annotation of the sample feature vector can be completed by professionals in advance.
[0077] Optionally, the sample environment attribute data and the sample machine attribute data in each sample scene data in the training set can be arranged based on the preset arrangement order, and then the sample feature vector corresponding to each sample scene data can be generated. Further, the corresponding label sequence number can be annotated for each sample feature vector according to the sample cleaning mode label annotated for each sample scene data. For example, if the sample cleaning mode label corresponding to a sample scene data is as shown in Table 1 above, the label sequence number 1 can be annotated for the sample feature vector of the sample scene data; if the sample cleaning mode label corresponding to a sample scene data is as shown in Table 2 below, the label sequence number 2 can be annotated for the sample feature vector of the sample scene data.
[0078] Table 2
[0079]
[0080] It can be understood that the sample scene data comprises the sample feature vector obtained by arranging the sample environment attribute data and the sample machine attribute data in a preset order, and the label sequence number of the sample cleaning mode label corresponding to the sample feature vector, which can make the sample scene data more orderly and concise, and further achieve the effect of more conveniently and efficiently training the to-be-trained model.
[0081] In an exemplary embodiment, the to-be-trained model can comprise a decision tree model. In the case where the to-be-trained model is a decision tree model, further refining the cleaning mode determination model obtained by training the to-be-trained model to convergence according to the plurality of sample scene data in the training set, please refer to the following description of the embodiment. Figure 2 and specifically comprising the following steps:
[0082] S201, determining a first information entropy according to a plurality of sample scene data in a training set, and determining a second information entropy corresponding to each attribute data according to sample environment attribute data and sample machine attribute data of the plurality of sample scene data.
[0083] The first information entropy can be used to represent the degree of confusion of the sample scene data in the training set caused by different sample cleaning mode labels; the second information entropy can be used to represent the degree of confusion of the sample scene data related to each attribute data in the sample scene data caused by different sample cleaning mode labels. The attribute data can be data corresponding to the attribute in the sample scene data, for example, the attribute data in a certain sample scene data can include good lighting, spring season, 30% air humidity, good air quality, etc.
[0084] Optionally, the plurality of sample scene data in the training set can be counted to determine the number of sample scene data corresponding to the label number of different sample cleaning mode labels in the sample scene data, and then the first information entropy can be determined by the following formula (1).
[0085]
[0086] Wherein, H(R) refers to the first information entropy; R refers to the sample scene data; n refers to the total number of sample scene data in the training set; C i refers to the number of sample scene data with the label number i of the sample cleaning mode label; p(C i ) refers to the probability of the sample scene data with the label number i of the sample cleaning mode label in the sample scene data, which can be calculated by the following formula (2).
[0087]
[0088] Further, each attribute data included in the sample environment attribute data and the sample machine attribute data corresponding to the label number of different sample cleaning mode labels in the sample scene data can be counted to determine the number of sample scene data corresponding to the label number of different sample cleaning mode labels of each attribute data in the sample scene data, and then the second information entropy can be determined by the following formula (3).
[0089]
[0090] wherein H(R|a) refers to the second information entropy of the attribute data being a; a j refers to the jth attribute data; m refers to the total number of attribute data in the sample scene data; p(C i |a=a j ) refers to the probability of the sample scene data in the sample scene data, the attribute data being a j sample scene data corresponding to the label sequence number i of the sample cleaning mode label; p(a=a j ) refers to the probability of the attribute data being the jth attribute data in the sample scene data.
[0091] S202, determining a classification priority of each attribute data according to the first information entropy and the second information entropy corresponding to each attribute data.
[0092] wherein the classification priority is the priority of each attribute data for determining the sample cleaning mode corresponding to the sample scene data.
[0093] Optionally, after determining the first information entropy and the second information entropy of each attribute data, the first information entropy and the second information entropy of each attribute data can be input into a pre-set sequence determination model, and the classification priority of the attribute data is determined according to the first information entropy and the second information entropy of each attribute data through the sequence determination model.
[0094] S203, obtaining the cleaning mode determination model based on the decision tree model according to the classification priority of each attribute data, the plurality of sample scene data and the sample cleaning mode label.
[0095] Optionally, in the case that the model to be trained is a decision tree model, the hierarchical relationship of each attribute data can be determined according to the classification priority of the attribute data, and then a decision tree model composed of each hierarchical attribute data is generated. For example, if the attribute data with the highest priority is the remaining power and the attribute data with the lowest priority is the air quality, the attribute data with the highest priority, i.e. the remaining power, can be used as the first layer attribute data for classification in the decision tree model, and the attribute data with the lowest priority, i.e. the air quality, can be used as the last layer attribute data for classification in the decision tree model. Further, after obtaining the decision tree model including the hierarchical relationship of each attribute data, the sample cleaning mode label corresponding to each path in the decision tree model is determined based on the decision tree model according to the plurality of sample scene data and the sample cleaning mode label corresponding to the sample scene data, and then the cleaning mode determination model is obtained.
[0096] It can be understood that by introducing the first information entropy determined according to the sample scene data and the second information entropy corresponding to each attribute data included in the sample environment attribute data and the sample machine attribute data in the sample scene data, the classification priority of each attribute data can be more intuitively and accurately determined according to the first information entropy and the second information entropy corresponding to each attribute data, and then a more efficient and accurate effect of obtaining the cleaning mode determination model can be realized according to the classification priority and the sample cleaning mode label.
[0097] On the basis of the above-mentioned embodiments, in one exemplary embodiment, please refer to the accompanying Figure 3 Further, the above S202 is refined, which can specifically include the following steps:
[0098] S301, determining the information amount corresponding to each attribute data according to the first information entropy and the second information entropy corresponding to each attribute data.
[0099] The information amount refers to the influence degree of each attribute data on different sample cleaning mode labels corresponding to the sample scene data.
[0100] Optionally, after determining the first information entropy and the second information entropy of each attribute data, the information amount corresponding to each attribute data can be determined based on a pre-set calculation logic according to the first information entropy and the second information entropy of each attribute data. For example, the difference between the first information entropy and each second information entropy can be determined by the following formula (4), and then the difference between the first information entropy and the second information entropy corresponding to each attribute data is taken as the information amount corresponding to each attribute data.
[0101] I(R, a) = H(R) - H(R|a) (4)
[0102] Wherein, I(R, a) refers to the information amount corresponding to the attribute data a.
[0103] It should be noted that by determining the difference between the first information entropy and each second information entropy, the difference is taken as the information amount corresponding to each attribute data, which can more conveniently and accurately obtain the information amount corresponding to each attribute data.
[0104] S302, arranging all attribute data in descending order according to the corresponding information amount to obtain the classification priority of each attribute data.
[0105] Optionally, after determining the information amount corresponding to each attribute data, the information amount corresponding to each attribute data can be arranged in descending order, and the determined sequence is taken as the classification priority of the attribute data.
[0106] It can be understood that, by arranging each information quantity in descending order according to the information quantity of each attribute data determined according to the first information entropy and the second information entropy of each attribute data, the classification priority of the attribute data can be more efficiently and accurately determined, and a more reasonable and accurate cleaning mode determination model is obtained.
[0107] The second aspect of the present disclosure provides a cleaning mode determination device. Please refer to the accompanying Figure 4 The cleaning mode determination device comprises:
[0108] The data acquisition module 401 is configured to acquire target scene data of a cleaning robot, wherein the target scene data comprises target environment attribute data of a to-be-cleaned area and target machine attribute data of the cleaning robot.
[0109] The mode determination module 402 is configured to input the target scene data into a cleaning mode determination model to obtain a target cleaning mode output by the cleaning mode determination model, wherein the target cleaning mode comprises target values of a plurality of cleaning parameters in the cleaning robot.
[0110] In one embodiment, the cleaning mode determination device further comprises:
[0111] The model training module is configured to train the cleaning mode determination model in the following manner:
[0112] The training model is trained to convergence according to a plurality of sample scene data in a training set to obtain the cleaning mode determination model, wherein the sample scene data comprises sample environment attribute data of a sample to-be-cleaned area and sample machine attribute data of the cleaning robot, the sample scene data is labeled with a sample cleaning mode label, and the sample cleaning mode label comprises sample values of a plurality of cleaning parameters in the cleaning robot.
[0113] In one embodiment, the sample scene data comprises a sample feature vector and a label serial number of a sample cleaning mode label corresponding to the sample feature vector, and the sample feature vector is obtained by arranging the sample environment attribute data and the sample machine attribute data in a preset order.
[0114] In one embodiment, the training model comprises a decision tree model.
[0115] In one embodiment, the model training module comprises:
[0116] The information entropy determination unit is configured to determine a first information entropy according to a plurality of sample scene data in a training set, and determine a second information entropy corresponding to each attribute data according to sample environment attribute data and sample machine attribute data of the plurality of sample scene data.
[0117] a priority determination unit, configured to determine a classification priority of each attribute data according to the first information entropy and a second information entropy corresponding to each attribute data;
[0118] a model determination unit, configured to obtain the cleaning mode determination model based on a decision tree model according to the classification priority of each attribute data, the plurality of sample scene data and the sample cleaning mode label.
[0119] In one of the embodiments, the priority determination unit can include:
[0120] an information amount determination sub-unit, configured to determine an information amount corresponding to each attribute data according to the first information entropy and the second information entropy corresponding to each attribute data;
[0121] a priority determination sub-unit, configured to arrange all attribute data in descending order according to the corresponding information amount to obtain the classification priority of each attribute data.
[0122] In one of the embodiments, the information amount determination sub-unit is specifically configured to:
[0123] determine a difference between the first information entropy and the second information entropy corresponding to each attribute data as the information amount corresponding to each attribute data.
[0124] In one of the embodiments, the target environment attribute data includes at least one of a weather state, an air state and a floor state of the area to be cleaned.
[0125] In one of the embodiments, the target machine attribute data includes at least one of power information, water information and consumable consumption information of the cleaning robot.
[0126] The implementation process of the functions and roles of each module in the above device is specifically described in the implementation process of the corresponding steps in the above method, which will not be repeated here.
[0127] The third aspect of the present disclosure provides a computer program product, including computer programs / instructions, which, when executed by a processor, implement the method of the first aspect.
[0128] For the apparatus embodiment and the computer program product embodiment, since they substantially correspond to the method embodiment, the relevant parts are referred to the part of the method embodiment. In addition, the apparatus embodiment described above is only illustrative, wherein the modules described as separate components can or can not be physically separated, and the components shown as modules can or can not be physical modules, i.e., can be located in one place or distributed on multiple network modules. Part or all of the modules can be selected to achieve the purpose of the present disclosure according to actual needs. Those skilled in the art can understand and implement without creative labor.
[0129] In a fourth aspect, the embodiment of the cleaning mode determination apparatus provided by the present disclosure can be applied to a computer device. Please refer to the accompanying drawings Figure 5 which shows an exemplary schematic diagram of the hardware of a computer device. For example, the device 500 can be a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, a fitness device, a personal digital assistant, etc.
[0130] The device 500 can include one or more of the following components: a processing component 501, a memory 502, a power supply component 503, a multimedia component 504, an audio component 505, an input / output (I / O) interface 506, a sensor component 507, and a communication component 508.
[0131] The processing component 501 usually controls overall operations of the device 500, such as operations associated with displaying, making phone calls, data communications, camera operations and recording operations. The processing component 501 can include one or more processors 509 to execute instructions to complete all or part of steps of the methods described above. In addition, the processing component 501 can include one or more modules to facilitate the interaction between the processing component 501 and other components. For example, the processing component 501 can include a multimedia module to facilitate the interaction between the multimedia component 504 and the processing component 501.
[0132] The memory 502 is configured to store various types of data to support operations of the device 500. Examples of these data include instructions for any application or method operating on the device 500, contact data, phonebook data, messages, pictures, videos, etc. The memory 502 can be implemented by any type of volatile or non-volatile storage devices or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk or optical disk.
[0133] Power component 503 provides power to the various components of device 500. Power component 503 can include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for device 500.
[0134] Multimedia component 504 includes a screen providing an output interface between device 500 and a user. In some embodiments, the screen includes a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen can be implemented as a touch screen to receive input signals from a user. The touch panel includes one or more touch sensors to sense touch, slide, and gesture on the touch panel. The touch sensors can not only sense a boundary of a touch or a slide action, but also detect duration and pressure related to the touch or slide operation. In some embodiments, multimedia component 504 includes a front camera and / or a rear camera. When device 500 is in an operation mode, such as a shooting mode or a video mode, the front camera and / or the rear camera can receive external multimedia data. Each of the front and rear cameras can be a fixed optical lens system or have a focal length and optical zoom capability.
[0135] Audio component 505 is configured to output and / or input audio signals. For example, audio component 505 includes a microphone (MIC) that is configured to receive external audio signals when device 500 is in an operation mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signals can be further stored in memory 502 or transmitted via communication component 508. In some embodiments, audio component 505 also includes a speaker for outputting audio signals.
[0136] I / O interface 506 provides an interface between processing component 501 and peripheral interface modules, which can be a keyboard, a click wheel, buttons, and the like. The buttons can include, but are not limited to, a home button, a volume button, a start button, and a lock button.
[0137] The sensor component 507 includes one or more sensors to provide the device 500 with state assessment of various aspects. For example, the sensor component 507 can detect the open / closed state of the device 500, relative positioning of components, such as a display and a keypad of the device 500, changes of position of the device 500 or a component of the device 500, presence or absence of user contact with the device 500, orientation or acceleration / deceleration / g-force and temperature changes of the device 500. The sensor component 507 can also include proximity sensor(s) configured to detect presence of nearby objects without any physical contact. The sensor component 507 can further include a light sensor(s) such as a CMOS or CCD image sensor for use in imaging applications. In some embodiments, the sensor component 507 can also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor or a temperature sensor.
[0138] The communication component 508 is configured to facilitate wired or wireless communication between the device 500 and another device. The device 500 can access a wireless network based on a communication standard, such as WiFi, 2G or 3G, 4G or 5G or a combination thereof. In an exemplary embodiment, the communication component 508 receives broadcast signals or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 508 further includes a Near Field Communication (NFC) module to facilitate short-range communication. For example, the NFC module can be implemented based on Radio Frequency Identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technology.
[0139] In an exemplary embodiment, the device 500 can be implemented with one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, micro-controllers, microprocessors or other electronic elements to perform the ABC method of the computer device.
[0140] In a fifth aspect, the present disclosure also provides, in an exemplary embodiment, a non-transitory computer-readable storage medium including instructions, such as the memory 502 including instructions, which can be executed by the processor 509 of the device 500 to perform the ABC method of the computer device. For example, the non-transitory computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disc, and an optical data storage device, etc.
[0141] The above describes particular embodiments of the present disclosure. Other embodiments are within the scope of the following claims. In some cases, acts or steps recited in the claims can be performed in a different order and still accomplish the desired results. Additionally, the processes depicted in the figures do not necessarily require the particular order shown or sequential order to achieve the desired results. In certain implementations, multitasking and parallel processing can be advantageous or necessary.
[0142] Other embodiments of the disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the features of the disclosure as set forth herein. It is intended that the disclosure be construed as including any paterns of embodiments and their equivalents following the principles of the disclosure and including modifications and improvements to the paterns of the disclosure as would occur to those skilled in the art to which the disclosure pertains. The specification and examples given herein are intended to be illustrative only and not in a limiting sense. The true scope of the present disclosure is set forth in the following claims.
[0143] It is to be understood that the present disclosure is not limited to the precise construction described and as shown in the attached figures, and that various modifications and changes can be made by those skilled in the art without departing from the scope of the present disclosure. The scope of the present disclosure is limited only by the claims that follow.
[0144] The above description is intended to be illustrative and not restrictive. Many other embodiments will be apparent to those of skill in the art upon reading and understanding the above description. The scope of the disclosure should, therefore, be determined with reference to the appended claims, along with the full scope of equivalents to which such claims are entitled.
Claims
1. A cleaning mode determination method characterized by, The method comprises: obtaining target scene data of a cleaning robot, wherein the target scene data comprises target environment attribute data of a to-be-cleaned area and target machine attribute data of the cleaning robot; inputting the target scene data into a cleaning mode determination model to obtain a target cleaning mode output by the cleaning mode determination model; wherein the target cleaning mode comprises target values of multiple cleaning parameters in the cleaning robot.
2. The method of claim 1, wherein, The cleaning mode determination model is obtained by training in the following manner: training a to-be-trained model to convergence according to multiple sample scene data in a training set to obtain the cleaning mode determination model, wherein the sample scene data comprises sample environment attribute data of a sample to-be-cleaned area and sample machine attribute data of the cleaning robot, and the sample scene data is labeled with a sample cleaning mode label, and the sample cleaning mode label comprises sample values of multiple cleaning parameters in the cleaning robot.
3. The method of claim 2, wherein, The sample scene data comprises a sample feature vector and a label number of a sample cleaning mode label corresponding to the sample feature vector, and the sample feature vector is obtained by arranging the sample environment attribute data and the sample machine attribute data in a preset order.
4. The method of claim 2, wherein, The to-be-trained model comprises a decision tree model.
5. The method according to claim 4, characterized in that The training of the to-be-trained model to convergence according to multiple sample scene data in a training set to obtain the cleaning mode determination model comprises: determining a first information entropy according to the multiple sample scene data in the training set, and determining a second information entropy corresponding to each attribute data according to sample environment attribute data and sample machine attribute data of the multiple sample scene data; determining a classification priority of each attribute data according to the first information entropy and the second information entropy corresponding to each attribute data; obtaining the cleaning mode determination model according to the classification priority of each attribute data and the multiple sample scene data based on a decision tree model.
6. The method of claim 5, wherein, The determination of the classification priority of each attribute data according to the first information entropy and the second information entropy corresponding to each attribute data comprises: determining an information amount corresponding to each attribute data according to the first information entropy and the second information entropy corresponding to each attribute data; arranging all attribute data in descending order according to the corresponding information amount to obtain the classification priority of each attribute data.
7. The method of claim 6, wherein, The determination of the information amount corresponding to each attribute data according to the first information entropy and the second information entropy corresponding to each attribute data comprises: taking a difference value between the first information entropy and the second information entropy corresponding to each attribute data as the information amount corresponding to each attribute data.
8. The method of claim 1, wherein, The target environment attribute data comprises at least one of a weather state, an air state and a floor state of the to-be-cleaned area.
9. The method of claim 1, wherein, The target machine attribute data comprises at least one of power information, water information and consumable loss information of the cleaning robot.
10. A cleaning mode determining apparatus characterized by comprising: The apparatus comprises: a data acquisition module configured to obtain target scene data of a cleaning robot, wherein the target scene data comprises target environment attribute data of a to-be-cleaned area and target machine attribute data of the cleaning robot; A mode determining module is configured to input the target scene data into a cleaning mode determining model to obtain a target cleaning mode output by the cleaning mode determining model, wherein the target cleaning mode comprises target values of a plurality of cleaning parameters in the cleaning robot.
11. A computer program product comprising computer programs / instructions, characterized in that, The computer program / instructions, when executed by a processor, implement the method of any one of claims 1-9.
12. A computer device, comprising: A computer program product comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the method of any one of claims 1-9 when executing the program.
13. A computer readable storage medium having stored thereon a computer program, characterized in that, The program, when executed by a processor, implements the method of any one of claims 1-9.