Grass trimmer control method and device

Abstract

The invention provides a grass trimmer control method and device, and relates to the technical field of grass trimmer control. The method comprises the following steps: firstly, determining an area to be mowed, and determining a plurality of candidate mowing machine control schemes matched with a corresponding target mowing machine; secondly, analyzing a control scheme matching degree of the candidate grass trimmer control scheme relative to the to-be-trimmed area under each grass trimming control tendency; and then, after the control scheme matching degrees of the plurality of candidate grass trimmer control schemes relative to the to-be-trimmed region under the grass trimming control tendencies are obtained, determining a corresponding target grass trimmer control scheme in the plurality of candidate grass trimmer control schemes based on the target grass trimming control tendency corresponding to the to-be-trimmed region. On the basis of the content, the problem that in the prior art, the control reliability of the grass trimmer is relatively low can be solved.

Description

Technical Field

[0001] The present application relates to the technical field of lawn mower control, and in particular to a lawn mower control method and device. Background Art

[0002] Lawn mowers have many applications in agriculture, forestry, gardening and other scenarios, which can greatly improve the efficiency of mowing and reduce labor costs. Among them, in order to further reduce the time cost and difficulty of using the lawn mower, in the prior art, the lawn mower is generally configured with a corresponding working mode, such as a corresponding fixed configuration according to general application scenarios when leaving the factory, or the user performs a corresponding configuration according to actual conditions during use. However, the inventors have found through research that the factory-based fixed configuration is difficult to apply to various scenarios. Therefore, there is a problem of low reliability in controlling the lawn mower. In addition, in the application scenario where the user configures the lawn mower himself, the user is limited by his own experience and other limitations, and it is difficult to ensure that the lawn mower can be reliably controlled. Summary of the invention

[0003] In view of this, the purpose of the present application is to provide a lawn mower control method and device to improve the problem of relatively low reliability of lawn mower control in the prior art.

[0004] To achieve the above objectives, this application adopts the following technical solutions: A lawn mower control method, comprising: Determine a to-be-mowed area, and determine a plurality of candidate mower control schemes adapted to a target mower corresponding to the to-be-mowed area, wherein the plurality of candidate mower control schemes are reference mower control schemes pre-configured for the target mower; Analyzing the matching degree of the candidate mower control scheme with respect to the to-be-mowed area under each mowing control tendency; After obtaining the control scheme matching degree of the multiple candidate lawn mower control schemes relative to the area to be lawn mowing under each lawn mowing control tendency, based on the target lawn mowing control tendency corresponding to the area to be lawn mowing, the target lawn mower control scheme corresponding to the area to be lawn mowing is determined among the multiple candidate lawn mower control schemes, wherein the target lawn mower control tendency is generated based on the selection operation performed by the lawn mowing management user of the area to be lawn mowing, and the target lawn mower control scheme is used to control the operation process of the target lawn mower.

[0005] In a preferred embodiment of the present application, in the above-mentioned mower control method, the step of analyzing the degree of matching of the candidate mower control scheme with respect to the mowing area under each mowing control tendency comprises: The control scheme description data corresponding to the candidate mower control scheme, the regional environment description data corresponding to the to-be-mowed area, and the correlation description data between the mower control scheme and the mowing area are loaded respectively, so that a plurality of different semantic mining models perform corresponding semantic mining to output a corresponding mowing scheme association vector set, wherein the correlation description data between the mower control scheme and the mowing area includes the historical regional environment description data and the historical control scheme description data corresponding to the to-be-mowed area, or includes the historical regional environment description data and the historical control scheme description data corresponding to the related mowing area, and the plurality of different semantic mining models include a first semantic mining model and second semantic mining models corresponding to each mowing control tendency; Performing at least one semantic association mining on each grass cutting scheme association vector set, and outputting corresponding grass cutting association mining vectors, wherein the grass cutting association mining vectors include at least one first grass cutting association mining vector and a second grass cutting association mining vector corresponding to each grass cutting control tendency; For each grass cutting control tendency, the following processing is performed respectively: based on the at least one first grass cutting associated mining vector and the second grass cutting associated mining vector corresponding to the grass cutting control tendency, the control scheme matching degree corresponding to the candidate grass cutting machine control scheme under the grass cutting control tendency is determined.

[0006] In a preferred embodiment of the present application, in the above-mentioned mower control method, the control scheme description data corresponding to the candidate mower control scheme, the regional environment description data corresponding to the mowing area, and the correlation description data between the mower control scheme and the mowing area are loaded respectively, so that multiple different semantic mining models perform corresponding semantic mining to output the corresponding mowing scheme association vector set, including: Using the data vectorization model in the control scheme matching network, vectorization processing is performed on the control scheme description data corresponding to the candidate mower control scheme, the regional environment description data corresponding to the mowing area, and the correlation description data between the mower control scheme and the mowing area, and the corresponding description data initial vector is output; Performing vector compression on each of the initial description data vectors, and outputting a description data compression vector corresponding to each of the initial description data vectors; Using the first semantic mining model in the control scheme matching network, semantic mining is performed on each of the initial vectors of the description data, and a corresponding grass cutting scheme association vector set is output; By using the second semantic mining models in the control scheme matching network, each of the description data compression vectors is semantically mined respectively, and a set of grass cutting scheme associated vectors corresponding to each of the second semantic mining models is output, wherein the first semantic mining model is a shared semantic mining model, and the second semantic mining model is a semantic mining model for the corresponding grass cutting control tendency.

[0007] In a preferred embodiment of the present application, in the above-mentioned lawn mower control method, each semantic mining model has a corresponding first parameter distribution and a second parameter distribution, and the first parameter distribution and the second parameter distribution of different semantic mining models are different; The step of using the first semantic mining model in the control scheme matching network to perform semantic mining on each of the initial vectors of the description data and outputting a corresponding grass cutting scheme association vector set includes: Respectively determine the local mean vectors corresponding to the local vectors in the description data initial vectors corresponding to the control scheme description data, the regional environment description data and the correlation description data; Using the first parameter distribution corresponding to the first semantic mining model, the local mean vector is vector compressed, and a local compressed vector corresponding to the local mean vector is output; Performing vector expansion on the local compression vector using a second parameter distribution corresponding to the first semantic mining model, and outputting a local expansion vector corresponding to the local compression vector; The local expansion vectors corresponding to the initial vectors of the description data are combined and processed to output a corresponding grass cutting scheme associated vector set.

[0008] In a preferred embodiment of the present application, in the above-mentioned grass trimmer control method, each of the grass trimming control tendencies further corresponds to a second association mining model; The step of performing at least one semantic association mining on each grass cutting scheme association vector set and outputting the corresponding grass cutting association mining vector comprises: Cascading and combining the grass cutting scheme associated vectors in a grass cutting scheme associated vector set mined by using the first semantic mining model, outputting the corresponding grass cutting scheme cascade vector, and performing multi-level deep mining and associated mining on the grass cutting scheme cascade vector using at least one first associated mining model in the control scheme matching network, and outputting the corresponding first grass cutting associated mining vector, wherein, in the process of performing multi-level deep mining and associated mining, the input vector of the deep mining and associated mining of the latter level includes the output vector of the deep mining and associated mining of the previous level, so that the mined first grass cutting associated mining vector includes high-level semantic information; A cascade combination is performed on each grass cutting scheme associated vector in a grass cutting scheme associated vector set mined by using each of the second semantic mining models, and a corresponding grass cutting scheme cascade vector is output; and multi-level deep mining and association mining are performed on the grass cutting scheme cascade vector using the corresponding second association mining model in the control scheme matching network, and a corresponding second grass cutting association mining vector is output.

[0009] In a preferred embodiment of the present application, in the above-mentioned mower control method, each of the mowing control tendencies further corresponds to a semantic merging mining model and a semantic matching output model, and the semantic mining parameter distributions corresponding to each two of the semantic merging mining models are different; The step of determining the control scheme matching degree corresponding to the candidate mower control scheme under the mowing control tendency based on the at least one first mowing associated mining vector and the second mowing associated mining vector corresponding to the mowing control tendency comprises: Loading the at least one first grass trimming associated mining vector and the second grass trimming associated mining vector to load into a corresponding semantic merging mining model in a control scheme matching network; Mapping the mining vector distribution formed by combining the at least one first grass cutting associated mining vector and the second grass cutting associated mining vector using the semantic mining parameter distribution corresponding to the semantic merging mining model, and outputting a grass cutting plan target vector corresponding to a grass cutting control tendency; The target vector of the mowing scheme is loaded to be loaded into the corresponding semantic matching output model in the control scheme matching network; The output function included in the semantic matching output model is used to determine the control scheme matching degree corresponding to the candidate mower control scheme under the mowing control tendency.

[0010] In a preferred embodiment of the present application, in the above-mentioned mower control method, the step of mapping the mining vector distribution formed by the combination of the at least one first mowing-related mining vector and the second mowing-related mining vector using the semantic mining parameter distribution corresponding to the semantic merging mining model, and outputting a mowing plan target vector corresponding to a corresponding mowing control tendency includes: According to the semantic mining parameter distribution corresponding to the semantic merging mining model and the corresponding mowing scheme cascade vector, the semantic merging parameter corresponding to the mowing control tendency is mined, wherein the mowing scheme cascade vector is formed by cascading a second mowing association mining vector corresponding to the mowing control tendency and the at least one first mowing association mining vector; The semantic merging parameter and the mowing scheme cascade vector are multiplied to output a mowing scheme target vector corresponding to the mowing control tendency.

[0011] In a preferred embodiment of the present application, in the above-mentioned mower control method, the step of analyzing the degree of matching of the candidate mower control scheme with respect to the mowing area under each mowing control tendency further includes: Determine a training grass cutting data cluster, wherein each training grass cutting data in the training grass cutting data cluster includes training control scheme description data corresponding to the training grass cutting machine control scheme, training area environment description data corresponding to the training grass cutting area, training correlation description data and actual matching data of the control scheme under each grass cutting control tendency; The candidate control scheme matching network is updated and optimized multiple times through the training grass cutting data cluster to form the control scheme matching network, wherein each update and optimization includes: Loading the training grass cutting data determined from the training grass cutting data cluster to the candidate control scheme matching network, and using the control scheme matching network to determine the control scheme matching degree corresponding to each grass cutting control tendency; According to the error between the control scheme matching degree corresponding to each grass cutting control tendency and the actual matching data of the control scheme, the parameters of the candidate control scheme matching network are updated.

[0012] In a preferred embodiment of the present application, in the above-mentioned mower control method, after obtaining the control scheme matching degree of the plurality of candidate mower control schemes relative to the to-be-mowed area under each mowing control tendency, the step of determining the target mower control scheme corresponding to the to-be-mowed area from among the plurality of candidate mower control schemes based on the target mowing control tendency corresponding to the to-be-mowed area includes: After obtaining the control scheme matching degrees of the multiple candidate mower control schemes with respect to the to-be-mowed area under each mowing control tendency, determining the control scheme matching degrees of the multiple candidate mower control schemes with respect to the to-be-mowed area under the target mowing control tendency based on the target mowing control tendency corresponding to the to-be-mowed area; Based on the control scheme matching degrees of the multiple candidate lawn mower control schemes relative to the area to be lawn mowing under the target lawn mowing control tendency, the candidate lawn mower control scheme corresponding to the control scheme matching degree with the maximum value is determined, and the candidate lawn mower control scheme is determined as the target lawn mower control scheme corresponding to the area to be lawn mowing.

[0013] The present application also provides a lawn mower control device, comprising: A scheme determination module, used for determining a to-be-mowed area, and determining a plurality of candidate mower control schemes adapted to a target mower corresponding to the to-be-mowed area, wherein the plurality of candidate mower control schemes are reference mower control schemes pre-configured for the target mower; A scheme matching module, used for analyzing the control scheme matching degree of the candidate mower control scheme with respect to the to-be-mowed area under each mowing control tendency; A scheme screening module is used to determine, after obtaining the control scheme matching degree of the multiple candidate mower control schemes relative to the area to be mowed under each mowing control tendency, a target mower control scheme corresponding to the area to be mowed among the multiple candidate mower control schemes based on the target mowing control tendency corresponding to the area to be mowed, wherein the target mower control tendency is generated based on the selection operation performed by a mowing management user of the area to be mowed, and the target mower control scheme is used to control the operation process of the target mower.

[0014] The present application provides a mower control method and device. First, determine the area to be mowed, and determine multiple candidate mower control schemes that are compatible with the corresponding target mower; second, analyze the control scheme matching degree of the candidate mower control scheme relative to the area to be mowed under each mowing control tendency; then, after obtaining the control scheme matching degree of multiple candidate mower control schemes relative to the area to be mowed under each mowing control tendency, determine the corresponding target mower control scheme from multiple candidate mower control schemes based on the target mowing control tendency corresponding to the area to be mowed. Based on the above content, since not only the control scheme matching degree of the candidate mower control scheme relative to the area to be mowed is determined, the target mower control scheme determined based on the control scheme matching degree is more matched with the actual situation of the area to be mowed, thereby improving the reliability of mower control. In addition, since the control scheme matching degree is analyzed in combination with the corresponding mowing control tendency, the accuracy of the obtained control scheme matching degree can be higher, thereby further improving the control reliability. Based on this, the problem of relatively low reliability of grass trimmer control in the prior art can be fully improved, so as to achieve reliable grass trimming (weeding) in the grass trimming area. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings.

[0016] Figure 1 A structural block diagram of an electronic device provided in an embodiment of the present application; Figure 2A schematic diagram of a flow chart of a lawn mower control method provided in an embodiment of the present application; Figure 3 An architectural diagram of a control scheme matching network provided in an embodiment of the present application; Figure 4 A block diagram of a lawn mower control device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.

[0018] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the present application for which protection is sought, but merely represents selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0019] like Figure 1 As shown, an embodiment of the present application provides an electronic device, wherein the electronic device may include a memory, a processor, and a lawn mower control device.

[0020] In detail, the memory and the processor are electrically connected directly or indirectly to achieve data transmission or interaction. For example, the memory and the processor can be electrically connected through one or more communication buses or signal lines. The lawn mower control device includes at least one software function module stored in the memory in the form of software or firmware. The processor is used to execute an executable computer program stored in the memory, for example, the software function module and computer program included in the lawn mower control device, so as to implement the lawn mower control method provided in the embodiment of the present application.

[0021] Optionally, the memory may be, but is not limited to, a random access memory (RAM), a read only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable read-only memory (EEPROM), etc.

[0022] Furthermore, the processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), a system on chip (SoC), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0023] Understandably, Figure 1 The structure shown is for illustration only, and the electronic device may also include Figure 1 More or fewer components as shown, or with Figure 1 The different configurations shown, for example, may also include a communication unit for exchanging information with other devices (such as a lawn mower).

[0024] Combination Figure 2 The present application also provides a lawn mower control method applicable to the electronic device. The steps defined in the process related to the lawn mower control method can be implemented by the electronic device. Figure 2 The specific process shown is explained in detail.

[0025] Step S110, determining a to-be-mowed area, and determining a plurality of candidate mower control schemes adapted to a target mower corresponding to the to-be-mowed area.

[0026] In an embodiment of the present application, the electronic device can determine a mowing area to be mowed, and determine a plurality of candidate mower control schemes adapted to a target mower corresponding to the mowing area to be mowed. The plurality of candidate mower control schemes are reference mower control schemes pre-configured for the target mower, and can be configured by the manufacturer corresponding to the target mower.

[0027] Step S120: Analyze the matching degree of the candidate mower control scheme with respect to the to-be-mowed area under each mowing control tendency.

[0028] In the embodiment of the present application, the electronic device can analyze the control scheme matching degree of the candidate mower control scheme relative to the mowing area under each mowing control tendency. That is to say, for a variety of different mowing areas (the types and quantities of objects in the area are different), a variety of candidate mower control schemes are configured, and these candidate mower control schemes are universal for the corresponding mowing areas. However, in actual application scenarios, there may be different mowing tendencies for the same mowing area (for example, in some cases, the mowing speed needs to be faster, in some cases, the mowing height needs to be lower, etc.), so it is necessary to analyze in combination with the mowing control tendency.

[0029] Step S130, after obtaining the control scheme matching degree of the multiple candidate mower control schemes relative to the area to be mowed under each mowing control tendency, based on the target mowing control tendency corresponding to the area to be mowed, determine the target mower control scheme corresponding to the area to be mowed from the multiple candidate mower control schemes.

[0030] In an embodiment of the present application, after obtaining the control scheme matching degree of the multiple candidate mower control schemes with respect to the to-be-mowed area under each mowing control tendency, the electronic device can determine the target mower control scheme corresponding to the to-be-mowed area from among the multiple candidate mower control schemes based on the target mowing control tendency corresponding to the to-be-mowed area. The target mower control tendency is generated based on the selection operation performed by the mowing management user of the to-be-mowed area, and the target mower control scheme is used to control the operation process of the target mower. Exemplarily, there are candidate mower control scheme 1, candidate mower control scheme 2 and candidate mower control scheme 3, and grass cutting control tendency a and grass cutting control tendency b. In this way, the control scheme matching degree A of candidate mower control scheme 1 relative to the area to be grass cut under grass cutting control tendency a, the control scheme matching degree B of candidate mower control scheme 2 relative to the area to be grass cut under grass cutting control tendency a, and the control scheme matching degree C of candidate mower control scheme 3 relative to the area to be grass cut under grass cutting control tendency a can be obtained, and the control scheme matching degree D of candidate mower control scheme 1 relative to the area to be grass cut under grass cutting control tendency b, the control scheme matching degree E of candidate mower control scheme 2 relative to the area to be grass cut under grass cutting control tendency b, and the control scheme matching degree F of candidate mower control scheme 3 relative to the area to be grass cut under grass cutting control tendency b can be obtained. Assuming that the target mowing control tendency is the mowing control tendency b and the control scheme matching degree F has a maximum value, the candidate mower control scheme 3 can be determined as the target mower control scheme corresponding to the to-be-mowed area.

[0031] Based on the above content, since not only the control scheme matching degree of the candidate mower control scheme relative to the area to be mowed is determined, the target mower control scheme determined based on the control scheme matching degree is more matched with the actual situation of the area to be mowed, thereby improving the reliability of the mower control. In addition, since the corresponding mowing control tendency is also combined when analyzing the control scheme matching degree, the accuracy of the obtained control scheme matching degree can be higher, thereby further improving the control reliability. Based on this, the problem of relatively low reliability of mower control existing in the prior art can be fully improved to achieve reliable mowing (weeding) of the area to be mowed.

[0032] It should be noted that for step S120, the specific method of analyzing the matching degree of the candidate mower control scheme with respect to the to-be-mowed area under each mowing control tendency is not limited and can be selected according to actual needs.

[0033] For example, in an alternative implementation, in order to improve the reliability of the matching degree of the analyzed control scheme, Figure 3 The above-mentioned step S120 may further include step S121, step S122 and step S123, and the specific content of each step is described as follows.

[0034] Step S121, respectively load the control scheme description data corresponding to the candidate mower control scheme, the regional environment description data corresponding to the mowing area, and the correlation description data between the mower control scheme and the mowing area, so that multiple different semantic mining models perform corresponding semantic mining to output the corresponding mowing scheme association vector set.

[0035] In an embodiment of the present application, the control scheme description data (which may be a text data) corresponding to the candidate mower control scheme, the regional environment description data (which may be a text data or an image data, i.e., obtained by collecting an image of the mowing area) corresponding to the mower control scheme and the mowing area, and the correlation description data between the mower control scheme and the mowing area may be loaded respectively, so that a plurality of different semantic mining models may perform corresponding semantic mining to output a corresponding mowing scheme association vector set. The correlation description data between the mower control scheme and the mowing area may include the historical regional environment description data and the historical control scheme description data corresponding to the mowing area, or may include the historical regional environment description data and the historical control scheme description data corresponding to the relevant mowing area, and the plurality of different semantic mining models may include a first semantic mining model and a second semantic mining model corresponding to each mowing control tendency. For example, the first semantic mining model performs semantic mining on the control scheme description data, the regional environment description data, and the related relationship description data, respectively, to obtain a corresponding set of grass cutting scheme association vectors (i.e., including the grass cutting scheme association vectors corresponding to the control scheme description data, the grass cutting scheme association vectors corresponding to the regional environment description data, and the grass cutting scheme association vectors corresponding to the related relationship description data). The first second semantic mining model performs semantic mining on the control scheme description data, the regional environment description data, and the related relationship description data, respectively, to obtain a corresponding set of grass cutting scheme association vectors. The second second semantic mining model performs semantic mining on the control scheme description data, the regional environment description data, and the related relationship description data, respectively, to obtain a corresponding set of grass cutting scheme association vectors.

[0036] Step S122, performing at least one semantic association mining on each grass cutting scheme association vector set, and outputting the corresponding grass cutting association mining vector.

[0037] In an embodiment of the present application, after obtaining the grass cutting scheme association vector set corresponding to each semantic mining model, semantic association mining can be performed at least once on each grass cutting scheme association vector set, and the corresponding grass cutting association mining vector can be output. Wherein, the grass cutting association mining vector includes at least one first grass cutting association mining vector and a second grass cutting association mining vector corresponding to each grass cutting control tendency. For example, semantic association mining is performed at least once on the grass cutting scheme association vector set corresponding to the first semantic mining model to obtain at least one corresponding first grass cutting association mining vector. Semantic association mining is performed on the grass cutting scheme association vector set corresponding to the first second semantic mining model to obtain a corresponding second grass cutting association mining vector. Semantic association mining is performed on the grass cutting scheme association vector set corresponding to the second second semantic mining model to obtain a corresponding second grass cutting association mining vector. The grass cutting control tendency corresponds one-to-one to the second semantic mining model, and the second semantic mining model can be trained for each grass cutting control tendency during the training process.

[0038] Step S123, for each grass cutting control tendency, perform the following processing respectively: determine the control scheme matching degree corresponding to the candidate grass cutting machine control scheme under the grass cutting control tendency according to the at least one first grass cutting associated mining vector and the second grass cutting associated mining vector corresponding to the grass cutting control tendency.

[0039] In an embodiment of the present application, after obtaining the grass cutting associated mining vector, the following processing can be performed for each grass cutting control tendency: based on the at least one first grass cutting associated mining vector and the second grass cutting associated mining vector corresponding to the grass cutting control tendency, determine the control scheme matching degree corresponding to the candidate grass cutting machine control scheme under the grass cutting control tendency. That is to say, the second grass cutting associated mining vector is unique to each grass cutting control tendency, and the first grass cutting associated mining vector is common to each grass cutting control tendency, that is, for grass cutting control tendency a, based on the second grass cutting associated mining vector corresponding to the grass cutting control tendency a described by at least one first grass cutting associated mining vector, determine the control scheme matching degree corresponding to the candidate grass cutting machine control scheme under the grass cutting control tendency a. For grass cutting control tendency b, based on the second grass cutting associated mining vector corresponding to the grass cutting control tendency b described by at least one first grass cutting associated mining vector, determine the control scheme matching degree corresponding to the candidate grass cutting machine control scheme under the grass cutting control tendency b.

[0040] It can be understood that, in the above-mentioned step S121, the specific method in which multiple different semantic mining models perform corresponding semantic mining to output the corresponding mowing plan associated vector set is not restricted and can be selected according to actual needs. For example, in an alternative implementation, in order to improve the reliability of semantic mining and ensure the accuracy of the semantic representation of the mined mowing plan associated vector set, the above-mentioned step S121 can further include step S121a, step S121b, step S121c and step S121d, and the specific content of each step is as follows.

[0041] Step S121a, using the data vectorization model in the control scheme matching network, respectively vectorizes the control scheme description data corresponding to the candidate lawn mower control scheme, the regional environment description data corresponding to the to-be-mowed area, and the correlation description data between the lawn mower control scheme and the lawn mowing area, and outputs the corresponding description data initial vector.

[0042] In an embodiment of the present application, the data vectorization model in the control scheme matching network can be used to vectorize the control scheme description data corresponding to the candidate mower control scheme, the regional environment description data corresponding to the mowing area, and the correlation description data between the mower control scheme and the mowing area, and output the corresponding description data initial vector, that is, to obtain the description data initial vector corresponding to the control scheme description data, the description data initial vector corresponding to the regional environment description data, and the description data corresponding to the correlation description data, wherein the text data can be vectorized based on the word embedding model included in the data vectorization model, and the image data can be vectorized based on the convolution model included in the data vectorization model. For example, for the control scheme description data "cutting speed is 3000RPM, cutting height is 2cm, forward speed is 3km / h", the corresponding description data initial vector can be: Cut: 0.1, −0.2, 0.3, 0.4, −0.1, 0.6, −0.3, 0.2, 0.5, −0.4, …; Speed: 0.2, 0.1, −0.4, 0.3, 0.0, 0.5, −0.2, 0.4, 0.1, −0.3, …; is: 0.0, −0.1, 0.2, 0.1, 0.2, −0.2, 0.3, −0.1, 0.0, 0.1, …; 3000RPM: 0.3, 0.4, −0.1, 0.0, −0.5, 0.2, 0.1, −0.3, 0.4, −0.2,…; Cut: 0.1, −0.2, 0.3, 0.4, −0.1, 0.6, −0.3, 0.2, 0.5, −0.4, …; Height: 0.4, −0.3, 0.1, 0.2, 0.0, 0.4, −0.1, 0.3, 0.5, −0.5, …; is: 0.0, −0.1, 0.2, 0.1, 0.2, −0.2, 0.3, −0.1, 0.0, 0.1, …; 2cm: 0.2, 0.3, 0.4, −0.3, 0.1, 0.2, 0.0, −0.4, 0.1, −0.3,…; forward: 0.3, 0.2, −0.4, 0.5, 0.1, −0.1, 0.2, 0.4, −0.2, 0.0, …; Speed: 0.2, 0.1, −0.4, 0.3, 0.0, 0.5, −0.2, 0.4, 0.1, −0.3, …; is: 0.0, −0.1, 0.2, 0.1, 0.2, −0.2, 0.3, −0.1, 0.0, 0.1, …; 3km / h: 0.5, 0.0, −0.2, 0.3, 0.4, −0.3, 0.1, 0.2, −0.1, 0.0, ..... Step S121b, performing vector compression on each of the initial description data vectors, and outputting a description data compression vector corresponding to each of the initial description data vectors.

[0043] In an embodiment of the present application, vector compression can be performed on each of the initial vectors of the description data, and a compressed description data vector corresponding to each of the initial vectors of the description data can be output, so that the dimension of the vector can be reduced, thereby reducing the complexity of subsequent processing. Specifically, compression can be achieved by performing pooling and other processing on the initial vectors of the description data.

[0044] Step S121c, using the first semantic mining model in the control scheme matching network, semantic mining is performed on each of the initial vectors of the description data, and a corresponding mowing scheme association vector set is output.

[0045] In an embodiment of the present application, the first semantic mining model in the control scheme matching network can be used to perform semantic mining on each of the initial vectors of the description data and output a corresponding set of mowing scheme association vectors, that is, by performing semantic mining on the initial vectors of the description data, the mowing scheme association vector in the set of mowing scheme association vectors can be obtained.

[0046] Step S121d, using each second semantic mining model in the control scheme matching network, respectively perform semantic mining on each of the description data compression vectors, and output a grass cutting scheme association vector set corresponding to each of the second semantic mining models.

[0047] In an embodiment of the present application, each second semantic mining model in the control scheme matching network can be used to perform semantic mining on each of the description data compression vectors, and output a set of grass cutting scheme association vectors corresponding to each of the second semantic mining models. Among them, the first semantic mining model belongs to a common semantic mining model, and the second semantic mining model belongs to a semantic mining model for the corresponding grass cutting control tendency. In other words, the first second semantic mining model can be used to perform semantic mining on each of the description data compression vectors, and output a set of grass cutting scheme association vectors corresponding to the first second semantic mining model. In addition, the second second semantic mining model can be used to perform semantic mining on each of the description data compression vectors, and output a set of grass cutting scheme association vectors corresponding to the second second semantic mining model.

[0048] It can be understood that, in the above step S121c, the specific manner of performing semantic mining on each of the initial vectors of the description data is not limited and can be selected according to actual needs. For example, in an alternative implementation, in order to improve the reliability of semantic mining, each semantic mining model has a corresponding first parameter distribution and a second parameter distribution, and different semantic mining models have different first parameter distributions and second parameter distributions. Based on this, the above step S121c can further include the following implementable content (similar to step S121d): First, the local mean vectors corresponding to the local vectors in the description data initial vectors corresponding to the control scheme description data, the regional environment description data and the correlation description data can be determined respectively; illustratively, for each row vector in the description data initial vector, the row vector is used as a local vector, and then the mean of each local vector is calculated to obtain the corresponding local mean vector, so that the local mean vector corresponding to each description data can be obtained; Secondly, the local mean vector can be vector compressed by using the first parameter distribution corresponding to the first semantic mining model, and a local compressed vector corresponding to the local mean vector can be output; illustratively, the size of the first parameter distribution can be consistent with the size of the local mean vector, so that the local mean vector and the first parameter distribution can be bitwise multiplied to obtain the corresponding local compressed vector, wherein the first parameter distribution is formed in the corresponding training process as the network parameter of the first semantic mining model, and each parameter in the first parameter distribution is greater than or equal to 0 and less than or equal to 1; Then, the local compression vector can be vector-expanded by using the second parameter distribution corresponding to the first semantic mining model, and the local expansion vector corresponding to the local compression vector can be output; illustratively, the size of the second parameter distribution can be consistent with the size of the local compression vector, so that the local compression vector and the second parameter distribution can be bitwise multiplied to obtain the corresponding local expansion vector, wherein the second parameter distribution is formed in the corresponding training process as the network parameter of the first semantic model, and each parameter in the second parameter distribution is greater than or equal to 1; Finally, the local expansion vectors corresponding to the initial vectors of the description data can be combined to output a corresponding mowing plan associated vector set, that is, the mowing plan associated vector set includes each local expansion vector, that is, each local expansion vector is used as a mowing plan associated vector.

[0049] It can be understood that, in the above step S122, semantic association mining is performed at least once for each grass cutting scheme association vector set, and the specific method of outputting the corresponding grass cutting association mining vector is not limited and can be selected according to actual needs. For example, in an alternative implementation, in order to ensure the reliability of semantic association mining, each grass cutting control tendency further corresponds to a second association mining model, and the above step S122 can further include the following contents: First, a cascade combination is performed on each grass cutting scheme associated vector in a grass cutting scheme associated vector set mined by using the first semantic mining model, and a corresponding grass cutting scheme cascade vector (i.e., {grass cutting scheme associated vector 1, grass cutting scheme associated vector 2, grass cutting scheme associated vector 3}, which can also be understood as vector concat) is output, and at least one first associated mining model in the control scheme matching network is used to perform multi-level deep mining and associated mining on the grass cutting scheme cascade vector, and a corresponding first grass cutting associated mining vector is output (wherein, one first associated mining model corresponds to one first grass cutting associated mining vector), wherein, in the process of performing multi-level deep mining and associated mining, the input vector of the deep mining and associated mining of the latter level includes the output vector of the deep mining and associated mining of the previous level, so that the mined first grass cutting associated mining vector includes high-level semantic information; Secondly, each grass cutting scheme associated vector in a grass cutting scheme associated vector set mined by using the second semantic mining model is cascaded and combined, and a corresponding grass cutting scheme cascade vector is output; and, using the corresponding second associated mining model in the control scheme matching network, the grass cutting scheme cascade vector is subjected to multi-level deep mining and association mining, and a corresponding second grass cutting association mining vector is output.

[0050] It can be understood that, in the above steps, the specific manner of performing multi-level deep mining and association mining on the mowing scheme cascade vector to output the corresponding mowing association mining vector is not limited. For example, in an alternative implementation, the following contents may be included: First, the mowing scheme cascade vector may be subjected to multiple convolution processes in sequence to obtain multiple corresponding convolution vectors, wherein the input vector of the first convolution process is the mowing scheme cascade vector, the input vector of the second convolution process is the convolution vector output by the first convolution process, the input vector of the third convolution process is the convolution vector output by the second convolution process, and so on; Secondly, each of the convolution vectors is pooled to obtain a corresponding pooled vector, wherein the pooling method corresponding to each convolution vector is different, so that the sizes of the obtained pooled vectors are the same, or, after the same pooling process, corresponding sampling process is performed to make the final pooled vectors have the same size; Then, the multiple pooling vectors can be combined in pairs to form corresponding multiple pooling vector combinations (each pooling vector combination is composed of a first pooling vector and a second pooling vector), and, for each of the pooling vector combinations, the product between the corresponding first pooling vector and the corresponding second pooling vector is calculated, and, based on the product, the second pooling vector is weighted to obtain a corresponding association vector; Finally, the multiple association vectors corresponding to the multiple pooling vector combinations can be averaged to obtain the corresponding grass cutting association mining vector.

[0051] It can be understood that, in the above-mentioned step S123, the specific method of determining the control scheme matching degree corresponding to the candidate mower control scheme under the mowing control tendency is not restricted and can be selected according to actual needs. For example, in an alternative embodiment, in order to improve the reliability of the determined control scheme matching degree, each of the mowing control tendencies also corresponds to a semantic merging mining model and a semantic matching output model, and the semantic mining parameter distribution corresponding to each two of the semantic merging mining models (which can be distributed in a matrix and can be formed in the corresponding training process as the network parameters of the semantic merging model) is different. Based on this, the above-mentioned step S123 can further include step S123a, step S123b, step S123c and step S123d, and the specific content of each step is described as follows.

[0052] Step S123a: loading the at least one first grass trimming associated mining vector and the second grass trimming associated mining vector to load into a corresponding semantic merging mining model in a control solution matching network.

[0053] In an embodiment of the present application, the at least one first grass cutting associated mining vector and the second grass cutting associated mining vector can be loaded to be loaded into the corresponding semantic merging mining model in the control scheme matching network, so that the semantic merging mining model can obtain the at least one first grass cutting associated mining vector and the second grass cutting associated mining vector.

[0054] Step S123b, using the semantic mining parameter distribution corresponding to the semantic merging mining model, maps the mining vector distribution formed by the combination of the at least one first grass cutting associated mining vector and the second grass cutting associated mining vector, and outputs a grass cutting plan target vector corresponding to a corresponding grass cutting control tendency.

[0055] In an embodiment of the present application, after obtaining the at least one first grass cutting associated mining vector and the second grass cutting associated mining vector, the semantic mining parameter distribution corresponding to the semantic merging mining model can be used to map the mining vector distribution formed by the combination of the at least one first grass cutting associated mining vector and the second grass cutting associated mining vector, and output a corresponding grass cutting plan target vector corresponding to the grass cutting control tendency.

[0056] Step S123c, loading the grass cutting scheme target vector to load into the corresponding semantic matching output model in the control scheme matching network.

[0057] In an embodiment of the present application, the mowing plan target vector may also be loaded to be loaded into a corresponding semantic matching output model in the control scheme matching network, so that the semantic matching output model may obtain the mowing plan target vector.

[0058] Step S123d: using the output function included in the semantic matching output model, determine the control scheme matching degree corresponding to the candidate mower control scheme under the mowing control tendency.

[0059] In the embodiment of the present application, after obtaining the target vector of the mowing scheme, the output function included in the semantic matching output model can be used to determine the control scheme matching degree corresponding to the candidate mower control scheme under the mowing control tendency. Exemplarily, the semantic matching output model can first perform full connection processing on the mowing scheme target vector to obtain the corresponding full connection vector, and then output the full connection vector based on the corresponding output function (such as softmax, etc.) to obtain the corresponding probability (which can be the probability distribution corresponding to each candidate mower control scheme).

[0060] It can be understood that, in the above step S123b, the specific manner of mapping the mining vector distribution formed by the combination of the at least one first grass cutting associated mining vector and the second grass cutting associated mining vector is not limited and can be selected according to actual needs. For example, in an alternative implementation, in order to improve the semantic representation reliability of the output grass cutting plan target vector, the above step S123b can further include the following contents: First, based on the semantic mining parameter distribution corresponding to the semantic merging mining model and the corresponding mowing scheme cascade vector, the semantic merging parameter corresponding to the mowing control tendency can be mined, wherein the mowing scheme cascade vector is formed by cascading the second mowing associated mining vector corresponding to the mowing control tendency and the at least one first mowing associated mining vector. Exemplarily, the semantic mining parameter distribution and the corresponding mowing scheme cascade vector can be multiplied, and the multiplied product can be normalized to obtain the corresponding semantic merging parameter; Secondly, the semantic merging parameter and the mowing plan cascade vector can be multiplied to output the mowing plan target vector corresponding to the mowing control tendency. That is to say, the semantic merging parameter can be used as a correlation characterization parameter between the semantic mining parameter distribution and the corresponding mowing plan cascade vector. In this way, the mowing plan cascade vector is weighted based on the semantic merging parameter to realize association mining and obtain the corresponding mowing plan target vector.

[0061] On the basis of the above content, it is also necessary to explain that step S120 may also include a step of training to form the control scheme matching network, which may specifically include: First, a training grass cutting data cluster can be determined, wherein each training grass cutting data in the training grass cutting data cluster includes training control scheme description data corresponding to the training grass cutting machine control scheme, training area environment description data corresponding to the training grass cutting area, training correlation description data, and actual matching data (i.e., label) of the control scheme under each grass cutting control tendency; Secondly, the candidate control scheme matching network is updated and optimized multiple times through the training grass cutting data cluster (until the optimization target is reached, such as error convergence, etc.) to form the control scheme matching network, wherein each update optimization includes: The first step is to load the training grass cutting data determined from the training grass cutting data cluster to the candidate control scheme matching network, and use the control scheme matching network to determine the control scheme matching degree corresponding to each grass cutting control tendency, and the specific process is as described above; In the second step, according to the error between the matching degree of the control scheme corresponding to each grass cutting control tendency and the actual matching data of the control scheme (the corresponding error can be calculated based on the cross entropy calculation formula), the parameters of the candidate control scheme matching network are updated, that is, each parameter is updated and adjusted in the direction of reducing the error.

[0062] It should be noted that for step S130 , the specific method of determining the target mower control scheme corresponding to the to-be-mowed area from among the plurality of candidate mower control schemes is not limited.

[0063] For example, in an alternative implementation, the above step S130 may further include the following detailed implementation contents: First, after obtaining the control scheme matching degrees of the multiple candidate mower control schemes with respect to the to-be-mowed area under each mowing control tendency, the control scheme matching degrees of the multiple candidate mower control schemes with respect to the to-be-mowed area under the target mowing control tendency corresponding to the to-be-mowed area may be determined; Secondly, based on the control scheme matching degrees of the multiple candidate lawn mower control schemes relative to the area to be lawn mowing under the target lawn mowing control tendency, the candidate lawn mower control scheme corresponding to the control scheme matching degree with the maximum value can be determined, and the candidate lawn mower control scheme can be determined as the target lawn mower control scheme corresponding to the area to be lawn mowing.

[0064] Combination Figure 4The embodiment of the present application further provides a lawn mower control device applicable to the above electronic device. The lawn mower control device may include a solution determination module, a solution matching module and a solution screening module.

[0065] The scheme determination module is used to determine the area to be mowed and determine multiple candidate mower control schemes suitable for the target mower corresponding to the area to be mowed, wherein the multiple candidate mower control schemes are reference mower control schemes pre-configured for the target mower. Figure 2 As shown in step S110, the relevant contents of the solution determination module can refer to the above description of step S110.

[0066] The scheme matching module is used to analyze the control scheme matching degree of the candidate mower control scheme relative to the to-be-mowed area under each mowing control tendency. In the embodiment of the present application, the scheme matching module can be used to execute Figure 2 As shown in step S120, the relevant contents of the solution matching module can refer to the above description of step S120.

[0067] The scheme screening module is used to determine the target mower control scheme corresponding to the to-be-mowed area from among the multiple candidate mower control schemes after obtaining the control scheme matching degree of the multiple candidate mower control schemes relative to the to-be-mowed area under each mowing control tendency, based on the target mowing control tendency corresponding to the to-be-mowed area, wherein the target mower control tendency is generated based on the selection operation performed by the mowing management user of the to-be-mowed area, and the target mower control scheme is used to control the operation process of the target mower. In the embodiment of the present application, the scheme screening module can be used to execute Figure 2 As shown in step S130, the relevant contents of the solution screening module can refer to the above description of step S130.

[0068] In an embodiment of the present application, corresponding to the above-mentioned lawn mower control method applied to the electronic device, a computer-readable storage medium is also provided, in which a computer program is stored, and when the computer program is run, each step of the lawn mower control method is executed.

[0069] Among them, the steps executed when the aforementioned computer program is running will not be described one by one here, and reference may be made to the explanation of the aforementioned lawn mower control method in the previous text.

[0070] In summary, the present application provides a lawn mower control method and device, first, determine the area to be mowed, and determine multiple candidate lawn mower control schemes adapted to the corresponding target lawn mower; second, analyze the control scheme matching degree of the candidate lawn mower control scheme relative to the area to be mowed under each mowing control tendency; then, after obtaining the control scheme matching degree of multiple candidate lawn mower control schemes relative to the area to be mowed under each mowing control tendency, based on the target mowing control tendency corresponding to the area to be mowed, determine the corresponding target lawn mower control scheme from multiple candidate lawn mower control schemes. Based on the above content, since not only the control scheme matching degree of the candidate lawn mower control scheme relative to the area to be mowed is determined, the target lawn mower control scheme determined based on the control scheme matching degree is more matched with the actual situation of the area to be mowed, thereby improving the reliability of lawn mower control. In addition, since the control scheme matching degree is analyzed in combination with the corresponding mowing control tendency, the accuracy of the obtained control scheme matching degree can be higher, thereby further improving the control reliability. Based on this, the problem of relatively low reliability of grass trimmer control in the prior art can be fully improved to achieve reliable grass trimming in the grass trimming area.

[0071] In several embodiments provided in the embodiments of the present application, it should be understood that the disclosed device and method can also be implemented in other ways. The device and method embodiments described above are merely schematic, for example, the flowcharts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the device, method and computer program product according to multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of a code contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0072] In addition, the functional modules in the various embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.

[0073] If the function is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, electronic device, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program code. It should be noted that in this article, the term "include", "include" or any other variant thereof is intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements includes not only those elements, but also includes other elements that are not explicitly listed, or also includes elements inherent to such process, method, article or device. Without more constraints, an element defined by the phrase "comprising a..." does not exclude the existence of other identical elements in the process, method, article or apparatus comprising the element.

[0074] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A lawn mower control method, characterized in that: include: Determine a to-be-mowed area, and determine a plurality of candidate mower control schemes adapted to a target mower corresponding to the to-be-mowed area, wherein the plurality of candidate mower control schemes are reference mower control schemes pre-configured for the target mower; Analyzing the matching degree of the candidate mower control scheme with respect to the to-be-mowed area under each mowing control tendency; After obtaining the control scheme matching degree of the multiple candidate lawn mower control schemes relative to the area to be lawn mowing under each lawn mowing control tendency, based on the target lawn mowing control tendency corresponding to the area to be lawn mowing, the target lawn mower control scheme corresponding to the area to be lawn mowing is determined among the multiple candidate lawn mower control schemes, wherein the target lawn mower control tendency is generated based on the selection operation performed by the lawn mowing management user of the area to be lawn mowing, and the target lawn mower control scheme is used to control the operation process of the target lawn mower.

2. The grass trimmer control method according to claim 1, characterized in that: The step of analyzing the degree of matching of the candidate mower control scheme with respect to the to-be-mowed area under each mowing control tendency comprises: The control scheme description data corresponding to the candidate mower control scheme, the regional environment description data corresponding to the to-be-mowed area, and the correlation description data between the mower control scheme and the mowing area are loaded respectively, so that a plurality of different semantic mining models perform corresponding semantic mining to output a corresponding mowing scheme association vector set, wherein the correlation description data between the mower control scheme and the mowing area includes the historical regional environment description data and the historical control scheme description data corresponding to the to-be-mowed area, or includes the historical regional environment description data and the historical control scheme description data corresponding to the related mowing area, and the plurality of different semantic mining models include a first semantic mining model and second semantic mining models corresponding to each mowing control tendency; Performing at least one semantic association mining on each grass cutting scheme association vector set, and outputting corresponding grass cutting association mining vectors, wherein the grass cutting association mining vectors include at least one first grass cutting association mining vector and a second grass cutting association mining vector corresponding to each grass cutting control tendency; For each grass cutting control tendency, the following processing is performed respectively: based on the at least one first grass cutting associated mining vector and the second grass cutting associated mining vector corresponding to the grass cutting control tendency, the control scheme matching degree corresponding to the candidate grass cutting machine control scheme under the grass cutting control tendency is determined.

3. The grass trimmer control method according to claim 2, characterized in that: The step of loading control scheme description data corresponding to the candidate mower control scheme, regional environment description data corresponding to the mowing area, and correlation description data between the mower control scheme and the mowing area respectively so that a plurality of different semantic mining models perform corresponding semantic mining to output a corresponding mowing scheme association vector set includes: Using the data vectorization model in the control scheme matching network, vectorization processing is performed on the control scheme description data corresponding to the candidate mower control scheme, the regional environment description data corresponding to the mowing area, and the correlation description data between the mower control scheme and the mowing area, and the corresponding description data initial vector is output; Performing vector compression on each of the initial description data vectors, and outputting a description data compression vector corresponding to each of the initial description data vectors; Using the first semantic mining model in the control scheme matching network, semantic mining is performed on each of the initial vectors of the description data, and a corresponding grass cutting scheme association vector set is output; By using the second semantic mining models in the control scheme matching network, each of the description data compression vectors is semantically mined respectively, and a set of grass cutting scheme associated vectors corresponding to each of the second semantic mining models is output, wherein the first semantic mining model is a shared semantic mining model, and the second semantic mining model is a semantic mining model for the corresponding grass cutting control tendency.

4. The grass trimmer control method according to claim 3, characterized in that: Each semantic mining model has a corresponding first parameter distribution and a second parameter distribution, and different semantic mining models have different first parameter distributions and second parameter distributions; The step of using the first semantic mining model in the control scheme matching network to perform semantic mining on each of the initial vectors of the description data and outputting a corresponding grass cutting scheme association vector set includes: Respectively determine the local mean vectors corresponding to the local vectors in the description data initial vectors corresponding to the control scheme description data, the regional environment description data and the correlation description data; Using the first parameter distribution corresponding to the first semantic mining model, the local mean vector is vector compressed, and a local compressed vector corresponding to the local mean vector is output; Performing vector expansion on the local compression vector using a second parameter distribution corresponding to the first semantic mining model, and outputting a local expansion vector corresponding to the local compression vector; The local expansion vectors corresponding to the initial vectors of the description data are combined and processed to output a corresponding grass cutting scheme associated vector set.

5. The grass trimmer control method according to claim 2, characterized in that: Each of the grass cutting control tendencies also corresponds to a second association mining model; The step of performing at least one semantic association mining on each grass cutting scheme association vector set and outputting the corresponding grass cutting association mining vector comprises: Cascading and combining the grass cutting scheme associated vectors in a grass cutting scheme associated vector set mined by using the first semantic mining model, outputting the corresponding grass cutting scheme cascade vector, and performing multi-level deep mining and associated mining on the grass cutting scheme cascade vector using at least one first associated mining model in the control scheme matching network, and outputting the corresponding first grass cutting associated mining vector, wherein, in the process of performing multi-level deep mining and associated mining, the input vector of the deep mining and associated mining of the latter level includes the output vector of the deep mining and associated mining of the previous level, so that the mined first grass cutting associated mining vector includes high-level semantic information; A cascade combination is performed on each grass cutting scheme associated vector in a grass cutting scheme associated vector set mined by using each of the second semantic mining models, and a corresponding grass cutting scheme cascade vector is output; and multi-level deep mining and association mining are performed on the grass cutting scheme cascade vector using the corresponding second association mining model in the control scheme matching network, and a corresponding second grass cutting association mining vector is output.

6. The grass trimmer control method according to claim 2, characterized in that: Each of the grass cutting control tendencies also corresponds to a semantic merging mining model and a semantic matching output model, and the semantic mining parameter distributions corresponding to each two of the semantic merging mining models are different; The step of determining the control scheme matching degree corresponding to the candidate mower control scheme under the mowing control tendency based on the at least one first mowing associated mining vector and the second mowing associated mining vector corresponding to the mowing control tendency comprises: Loading the at least one first grass trimming associated mining vector and the second grass trimming associated mining vector to load into a corresponding semantic merging mining model in a control scheme matching network; Mapping the mining vector distribution formed by combining the at least one first grass cutting associated mining vector and the second grass cutting associated mining vector using the semantic mining parameter distribution corresponding to the semantic merging mining model, and outputting a grass cutting plan target vector corresponding to a grass cutting control tendency; The target vector of the mowing scheme is loaded to be loaded into the corresponding semantic matching output model in the control scheme matching network; The output function included in the semantic matching output model is used to determine the control scheme matching degree corresponding to the candidate mower control scheme under the mowing control tendency.

7. The grass trimmer control method according to claim 6, characterized in that: The step of mapping the mining vector distribution formed by combining the at least one first grass cutting associated mining vector and the second grass cutting associated mining vector by using the semantic mining parameter distribution corresponding to the semantic merging mining model, and outputting a grass cutting plan target vector corresponding to a corresponding grass cutting control tendency comprises: According to the semantic mining parameter distribution corresponding to the semantic merging mining model and the corresponding mowing scheme cascade vector, the semantic merging parameter corresponding to the mowing control tendency is mined, wherein the mowing scheme cascade vector is formed by cascading a second mowing association mining vector corresponding to the mowing control tendency and the at least one first mowing association mining vector; The semantic merging parameter and the mowing scheme cascade vector are multiplied to output a mowing scheme target vector corresponding to the mowing control tendency.

8. The grass trimmer control method according to any one of claims 3 to 7, characterized in that: The step of analyzing the degree of matching of the candidate mower control scheme with respect to the to-be-mowed area under each mowing control tendency also includes: Determine a training grass cutting data cluster, wherein each training grass cutting data in the training grass cutting data cluster includes training control scheme description data corresponding to the training grass cutting machine control scheme, training area environment description data corresponding to the training grass cutting area, training correlation description data and actual matching data of the control scheme under each grass cutting control tendency; The candidate control scheme matching network is updated and optimized multiple times through the training grass cutting data cluster to form the control scheme matching network, wherein each update and optimization includes: Loading the training grass cutting data determined from the training grass cutting data cluster to the candidate control scheme matching network, and using the control scheme matching network to determine the control scheme matching degree corresponding to each grass cutting control tendency; According to the error between the control scheme matching degree corresponding to each grass cutting control tendency and the actual matching data of the control scheme, the parameters of the candidate control scheme matching network are updated.

9. The grass trimmer control method according to claim 1, characterized in that: The step of determining a target mower control scheme corresponding to the to-be-mowed area from among the multiple candidate mower control schemes based on the target mowing control tendency corresponding to the to-be-mowed area after obtaining the control scheme matching degrees of the multiple candidate mower control schemes with respect to the to-be-mowed area under each mowing control tendency comprises: After obtaining the control scheme matching degrees of the multiple candidate mower control schemes with respect to the to-be-mowed area under each mowing control tendency, determining the control scheme matching degrees of the multiple candidate mower control schemes with respect to the to-be-mowed area under the target mowing control tendency based on the target mowing control tendency corresponding to the to-be-mowed area; Based on the control scheme matching degrees of the multiple candidate lawn mower control schemes relative to the area to be lawn mowing under the target lawn mowing control tendency, the candidate lawn mower control scheme corresponding to the control scheme matching degree with the maximum value is determined, and the candidate lawn mower control scheme is determined as the target lawn mower control scheme corresponding to the area to be lawn mowing.

10. A lawn mower control device, characterized in that: include: A scheme determination module, used for determining a to-be-mowed area, and determining a plurality of candidate mower control schemes adapted to a target mower corresponding to the to-be-mowed area, wherein the plurality of candidate mower control schemes are reference mower control schemes pre-configured for the target mower; A scheme matching module, used for analyzing the control scheme matching degree of the candidate mower control scheme with respect to the to-be-mowed area under each mowing control tendency; A scheme screening module is used to determine, after obtaining the control scheme matching degree of the multiple candidate mower control schemes relative to the area to be mowed under each mowing control tendency, a target mower control scheme corresponding to the area to be mowed among the multiple candidate mower control schemes based on the target mowing control tendency corresponding to the area to be mowed, wherein the target mower control tendency is generated based on the selection operation performed by a mowing management user of the area to be mowed, and the target mower control scheme is used to control the operation process of the target mower.

Images