A method and system for determining device characteristics of smart home

Abstract

The present invention provides a method and system for determining device features of a smart home, which relate to the field of data processing technology. In the present invention, based on the reference operation behavior description vector corresponding to the reference smart home device, multiple reference smart home devices are classified to form at least one reference device set; the type device behavior description vector corresponding to each reference device set is analyzed; based on the type device behavior description vector and the operation behavior description vector corresponding to the smart home device to be determined, the reference device set that best matches the smart home device to be determined is analyzed, and the reference device set is marked as the corresponding matching reference device set; the device feature type information corresponding to the matching reference device set is marked to be marked as the device feature type information corresponding to the smart home device to be determined. Based on the above method, the reliability of device feature type determination can be improved.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method and system for determining device characteristics of a smart home. Background Art

[0002] Smart home is reflected through the Internet of Things. Smart home connects various devices in the home (such as audio and video equipment, lighting system, curtain control, air conditioning control, security system, digital cinema system, network home appliances and three-meter copy, etc.) through the Internet of Things technology, providing home appliance control, lighting control, curtain control, telephone remote control, indoor and outdoor remote control, anti-theft alarm, environmental monitoring, HVAC control, infrared forwarding and programmable timing control and other functions and means. Compared with ordinary homes, smart homes not only have traditional residential functions, but also have architecture, network communication, information home appliances, equipment automation, and integrate systems, structures, services, and management into an efficient, comfortable, safe, convenient, and environmentally friendly living environment. It provides a full range of information interaction functions, helps families maintain smooth information exchanges with the outside world, optimizes people's lifestyles, helps people effectively arrange time, enhances the safety of home life, and even saves money for various energy costs.

[0003] Among them, in the application of smart home, based on some needs, it is necessary to analyze the smart home devices to determine the device feature type information corresponding to the smart home devices, that is, the device classification identification. However, in the prior art, there is a problem of low reliability in determining the device feature type. Summary of the invention

[0004] In view of this, an object of the present invention is to provide a method and system for determining device characteristics of a smart home, so as to improve the reliability of determining the type of device characteristics.

[0005] To achieve the above objectives, the embodiments of the present invention adopt the following technical solutions:

[0006] A method for determining characteristics of a smart home device, comprising:

[0007] Mining out the operation behavior description vector corresponding to the to-be-determined smart home device, and mining out the reference operation behavior description vector corresponding to the reference smart home device;

[0008] According to the reference operation behavior description vector corresponding to the reference smart home device, a plurality of the reference smart home devices are classified to form at least one corresponding reference device set, wherein any one of the at least one reference device set includes at least one reference smart home device having the same device feature type information;

[0009] Analyze the device behavior description vector corresponding to each of the reference device sets based on the reference operation behavior description vectors corresponding to the reference smart home devices.

[0010] According to the device behavior description vector of the category and the operation behavior description vector corresponding to the smart home device to be determined, in the at least one reference device set, a reference device set that best matches the smart home device to be determined is analyzed, and the reference device set is marked as a corresponding matching reference device set;

[0011] The device feature category information corresponding to the matching reference device set is marked as the device feature category information corresponding to the to-be-determined smart home device.

[0012] In some preferred embodiments, in the above-mentioned method for determining device features of smart home, before the step of mining out the operation behavior description vector corresponding to the smart home device to be determined and mining out the reference operation behavior description vector corresponding to the reference smart home device, the method for determining device features of smart home further includes:

[0013] Collect multi-angle device operation behavior data corresponding to the reference smart home device, and determine device feature type information corresponding to the reference smart home device;

[0014] Perform key information mining on the multi-angle device operation behavior data corresponding to the reference smart home device, and output an operation behavior description vector corresponding to the reference smart home device;

[0015] The initial neural network is updated by using typical device operation behavior data to output a corresponding candidate neural network, wherein the typical device operation behavior data includes an operation behavior description vector corresponding to the reference smart home device and corresponding device feature type information;

[0016] Performing a network reliability analysis operation on the candidate neural network to output a network reliability characterization parameter corresponding to the candidate neural network;

[0017] In the case that the network reliability characterization parameter is greater than or equal to the pre-configured reference reliability characterization parameter, the candidate neural network is marked as the corresponding operation behavior analysis neural network.

[0018] In some preferred embodiments, in the above-mentioned method for determining device features of a smart home, the step of analyzing the reference device set that best matches the smart home device to be determined from the at least one reference device set based on the device behavior description vector of the type and the operation behavior description vector corresponding to the smart home device to be determined, and marking the reference device set as the corresponding matching reference device set includes:

[0019] According to the corresponding device behavior description vector of the type and the operation behavior description vector of the to-be-determined smart home device, a reference device set that best matches the to-be-determined smart home device is analyzed from the at least one reference device set through the operation behavior analysis neural network;

[0020] The best matching reference device set is marked as the corresponding matching reference device set.

[0021] In some preferred embodiments, in the method for determining the device characteristics of the smart home, the step of analyzing the reference device set that best matches the smart home device to be determined from the at least one reference device set through the operation behavior analysis neural network based on the corresponding device behavior description vector of the type and the operation behavior description vector of the smart home device to be determined, comprises:

[0022] By using the operation behavior analysis neural network, according to the device behavior description vector of the type, analyzing the self-focus feature analysis results corresponding to the reference device set and the associated focus feature analysis results corresponding to the reference device set;

[0023] Analyze the vector matching parameters between the device behavior description vector of the type and the operation behavior description vector corresponding to the smart home device to be determined based on the associated focus feature analysis results and the self-focus feature analysis results corresponding to the device behavior description vector of the type;

[0024] Based on the vector matching parameter, a reference device set that best matches the to-be-determined smart home device is analyzed from the at least one reference device set.

[0025] In some preferred embodiments, in the above-mentioned method for determining device features of smart homes, the step of analyzing the self-focus feature analysis results and the corresponding associated focus feature analysis results corresponding to the reference device set according to the device behavior description vector by running the behavior analysis neural network includes:

[0026] By using the operation behavior analysis neural network, according to the device behavior description vector of the type corresponding to the reference device set, analyzing the self-focus feature analysis result;

[0027] Through the operation behavior analysis neural network, based on the operation behavior description vector corresponding to the smart home device to be determined and each of the device behavior description vectors corresponding to the reference device set, the associated focus feature analysis results of each of the reference device sets are analyzed.

[0028] In some preferred embodiments, in the above-mentioned method for determining device features of smart home, the step of analyzing the associated focus feature analysis results of each reference device set according to the operation behavior description vector corresponding to the to-be-determined smart home device and each of the device behavior description vectors of the type corresponding to the reference device set through the operation behavior analysis neural network includes:

[0029] Based on the first parameter distribution included in the operation behavior analysis neural network, weighted processing is performed on the operation behavior description vector corresponding to the to-be-determined smart home device to output the corresponding weighted operation behavior description vector, and, for each of the reference device sets, a row-column swap operation of vector parameters is performed on the type device behavior description vector corresponding to the reference device set to output the corresponding swapped device behavior description vector;

[0030] For each of the reference device sets, the exchange device behavior description vector corresponding to the reference device set and the weighted operation behavior description vector corresponding to the smart home device to be determined are multiplied, and the result of the multiplication is subjected to excitation mapping processing to output the excitation mapping description vector corresponding to the reference device set, and the category device behavior description vector corresponding to the reference device set and the excitation mapping description vector are multiplied to output the associated focus feature analysis result corresponding to the reference device set.

[0031] In some preferred embodiments, in the above-mentioned method for determining device features of smart homes, the step of analyzing the self-focus feature analysis result by running the behavior analysis neural network according to the type device behavior description vector corresponding to the reference device set includes:

[0032] Taking any one of the reference device sets as a first reference device set;

[0033] Based on the second parameter distribution matrix included in the operation behavior analysis neural network, weighting the category device behavior description vectors corresponding to the first reference device set to output corresponding weighted category behavior description vectors;

[0034] Performing a row-column swap operation on vector parameters of the category device behavior description vector corresponding to the first reference device set to output a corresponding swapped category behavior description vector;

[0035] Performing a first stimulus mapping process on the weighted type behavior description vector to output a first stimulus mapping description vector corresponding to the first reference device set, and performing a multiplication process on the swap type behavior description vector and the first stimulus mapping description vector, and performing a second stimulus mapping process on the result of the multiplication process to output a corresponding second stimulus mapping description vector, wherein the first stimulus mapping process is different from the second stimulus mapping process;

[0036] The category device behavior description vector corresponding to the first reference device set and the corresponding second excitation mapping description vector are multiplied to output the self-focus feature analysis result corresponding to the first reference device set.

[0037] In some preferred embodiments, in the above-mentioned method for determining the characteristics of smart home devices, the step of mining the operation behavior description vector corresponding to the smart home device to be determined includes:

[0038] Collecting multi-angle device operation behavior data of the to-be-determined smart home device;

[0039] Performing a key information mining operation on the multi-angle device operation behavior data corresponding to the to-be-determined smart home device to output operation behavior description vectors of the to-be-determined smart home device at multiple angles;

[0040] The operation behavior description vectors of multiple angles corresponding to the to-be-determined smart home device are subjected to vector cascade combination processing to output the operation behavior description vector corresponding to the to-be-determined smart home device.

[0041] The embodiment of the present invention further provides a system for determining characteristics of smart home devices, including:

[0042] A description vector mining module is used to mine the operation behavior description vector corresponding to the to-be-determined smart home device, and mine the reference operation behavior description vector corresponding to the reference smart home device;

[0043] A device classification module, configured to classify a plurality of the reference smart home devices according to the reference operation behavior description vectors corresponding to the reference smart home devices, so as to form at least one corresponding reference device set, wherein any one of the at least one reference device set includes at least one reference smart home device having the same device feature type information;

[0044] A category vector determination module, configured to analyze a category device behavior description vector corresponding to each reference device set based on a reference operation behavior description vector corresponding to the reference smart home device;

[0045] a matching set determination module, configured to analyze, based on the device behavior description vector of the category and the operation behavior description vector corresponding to the smart home device to be determined, a reference device set that best matches the smart home device to be determined in the at least one reference device set, and mark the reference device set as a corresponding matching reference device set;

[0046] The device feature type determination module is used to mark the device feature type information corresponding to the matching reference device set to mark it as the device feature type information corresponding to the smart home device to be determined.

[0047] In some preferred embodiments, in the above-mentioned smart home device feature determination system, the device feature determination system further includes at least one other functional module for:

[0048] Collect multi-angle device operation behavior data corresponding to the reference smart home device, and determine device feature type information corresponding to the reference smart home device;

[0049] Perform key information mining on the multi-angle device operation behavior data corresponding to the reference smart home device, and output an operation behavior description vector corresponding to the reference smart home device;

[0050] The initial neural network is updated by using typical device operation behavior data to output a corresponding candidate neural network, wherein the typical device operation behavior data includes an operation behavior description vector corresponding to the reference smart home device and corresponding device feature type information;

[0051] Performing a network reliability analysis operation on the candidate neural network to output a network reliability characterization parameter corresponding to the candidate neural network;

[0052] In the case that the network reliability characterization parameter is greater than or equal to the pre-configured reference reliability characterization parameter, the candidate neural network is marked as the corresponding operation behavior analysis neural network.

[0053] A method and system for determining device features of a smart home provided by an embodiment of the present invention can classify multiple reference smart home devices according to reference operation behavior description vectors corresponding to the reference smart home devices to form at least one reference device set; analyze the type device behavior description vector corresponding to each reference device set; analyze the reference device set that best matches the smart home device to be determined according to the type device behavior description vector and the operation behavior description vector corresponding to the smart home device to be determined, and mark the reference device set as the corresponding matching reference device set; mark the device feature type information corresponding to the matching reference device set to mark it as the device feature type information corresponding to the smart home device to be determined. Based on the foregoing, on the one hand, since the matching analysis is performed on the type device behavior description vector corresponding to the reference device set, the basis is more sufficient to improve the reliability of device feature type determination. In addition, the number of matching analyses can be reduced to make it more efficient.

[0054] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 A structural block diagram of a smart home device feature determination platform provided in an embodiment of the present invention.

[0056] Figure 2 A flowchart of the steps of a method for determining device characteristics of a smart home provided by an embodiment of the present invention.

[0057] Figure 3 A schematic diagram of various modules included in the smart home device feature determination system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0058] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0059] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0060] like Figure 1 As shown, an embodiment of the present invention provides a device feature determination platform for a smart home, wherein the device feature determination platform for a smart home may include a memory and a processor.

[0061] In detail, the memory and the processor are electrically connected directly or indirectly to realize data transmission or interaction. For example, they can be electrically connected to each other through one or more communication buses or signal lines. The memory may store at least one software function module (computer program) that may exist in the form of software or firmware. The processor may be used to execute the executable computer program stored in the memory, thereby realizing the device feature determination method of the smart home provided by the embodiment of the present invention (as described later).

[0062] Optionally, in some embodiments, the memory may be, but is not limited to, random access memory (RAM), read only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable read-only memory (Electric Erasable Programmable Read-Only Memory, EEPROM), etc.

[0063] Optionally, in some embodiments, 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.

[0064] Optionally, in some implementations, the smart home device feature determination platform may be a server with data processing capabilities.

[0065] Combination Figure 2The embodiment of the present invention further provides a method for determining device characteristics of a smart home, which can be applied to the device characteristics determination platform of the smart home. The method steps defined in the process related to the method for determining device characteristics of a smart home can be implemented by the device characteristics determination platform of the smart home. Figure 2 The specific process shown is explained in detail.

[0066] Step S110, mining the operation behavior description vector corresponding to the to-be-determined smart home device, and mining the reference operation behavior description vector corresponding to the reference smart home device.

[0067] In an embodiment of the present invention, the smart home device feature determination platform can mine out the operation behavior description vector corresponding to the smart home device to be determined, and mine out the reference operation behavior description vector corresponding to the reference smart home device.

[0068] Step S120: classify the plurality of reference smart home devices according to the reference operation behavior description vectors corresponding to the reference smart home devices to form at least one corresponding reference device set.

[0069] In an embodiment of the present invention, the device feature determination platform of the smart home can classify the plurality of reference smart home devices according to the reference operation behavior description vector corresponding to the reference smart home device, which can be implemented based on any clustering algorithm to form at least one corresponding reference device set. Any one of the at least one reference device set includes at least one reference smart home device having the same device feature type information.

[0070] Step S130: Analyze the device behavior description vectors corresponding to each of the reference device sets based on the reference operation behavior description vectors corresponding to the reference smart home devices.

[0071] In an embodiment of the present invention, the device feature determination platform of the smart home can analyze the type device behavior description vector corresponding to each of the reference device sets based on the reference operating behavior description vectors corresponding to the reference smart home devices. Exemplarily, for each of the reference device sets, the reference operating behavior description vectors corresponding to each reference smart home device included in the reference device set can be mean calculated to obtain the type device behavior description vector corresponding to the reference device set, or the cluster center of the reference operating behavior description vector corresponding to each reference smart home device included in the reference device set, that is, the analysis center for classifying multiple reference smart home devices based on the corresponding reference operating behavior description vectors, can be used as the type device behavior description vector corresponding to the reference device set.

[0072] Step S140, based on the device behavior description vector of the category and the operation behavior description vector corresponding to the smart home device to be determined, analyze the reference device set that best matches the smart home device to be determined in the at least one reference device set, and mark the reference device set as the corresponding matching reference device set.

[0073] In an embodiment of the present invention, the smart home device feature determination platform can analyze the reference device set that best matches the smart home device to be determined in the at least one reference device set based on the type device behavior description vector and the operation behavior description vector corresponding to the smart home device to be determined, and mark the reference device set as the corresponding matching reference device set.

[0074] Step S150: marking the device feature type information corresponding to the matching reference device set to mark it as the device feature type information corresponding to the to-be-determined smart home device.

[0075] In an embodiment of the present invention, the device feature determination platform of the smart home may mark the device feature category information corresponding to the matching reference device set to mark it as the device feature category information corresponding to the smart home device to be determined. The device feature category information may have different dimensions of content in different applications, or may not have any actual meaning and is only used to distinguish different smart home devices, such as codes 001, 002, 003, etc.

[0076] Based on the foregoing, on the one hand, since the matching analysis is performed on the device behavior description vector corresponding to the reference device set, the basis is more sufficient to improve the reliability of determining the device feature type. On the other hand, the number of matching analyses can be reduced, making it more efficient.

[0077] Optionally, in some implementations, the above step S110, i.e., the step of mining the operation behavior description vector corresponding to the to-be-determined smart home device (wherein, this step may be the same as or different from the step of mining the reference operation behavior description vector corresponding to the reference smart home device), may further include the following contents:

[0078] Collecting multi-angle device operation behavior data of the to-be-determined smart home device (such as data reflecting internal operation, data reflecting interaction with other devices, etc.);

[0079] Performing key information mining operations on the multi-angle device operation behavior data corresponding to the to-be-determined smart home device, such as mapping and convolution operations in feature space, to output operation behavior description vectors of the to-be-determined smart home device at multiple angles;

[0080] The operation behavior description vectors of multiple angles corresponding to the smart home device to be determined are subjected to vector cascade combination processing, such as combining them together to form a large description vector, so as to output the operation behavior description vector corresponding to the smart home device to be determined.

[0081] Optionally, in some implementations, before the step of mining out the operation behavior description vector corresponding to the to-be-determined smart home device and mining out the reference operation behavior description vector corresponding to the reference smart home device, that is, before step S110, the method for determining the device characteristics of the smart home may further include the following contents:

[0082] Collect multi-angle device operation behavior data corresponding to the reference smart home device, and determine device feature type information corresponding to the reference smart home device;

[0083] Perform key information mining on the multi-angle device operation behavior data corresponding to the reference smart home device, and output an operation behavior description vector corresponding to the reference smart home device;

[0084] The initial neural network is updated through typical device operation behavior data to output a corresponding candidate neural network, wherein the typical device operation behavior data includes an operation behavior description vector and corresponding device feature type information corresponding to the reference smart home device. For example, the corresponding predicted device feature type information can be predicted through the operation behavior description vector, and then a difference analysis is performed based on the predicted device feature type information and the device feature type information, and then the parameters of the initial neural network are updated based on the analyzed difference;

[0085] Performing a network reliability analysis operation on the candidate neural network to output a network reliability characterization parameter corresponding to the candidate neural network. Exemplarily, the network reliability analysis operation can be performed on the candidate neural network by configuring some detection device operation behavior data. The detection device operation behavior data can include an operation behavior description vector and device feature type information. In this way, the corresponding network reliability characterization parameter can be determined based on the accuracy between the predicted device feature type information and the actual device feature type information.

[0086] When the network reliability characterization parameter is greater than or equal to a pre-configured reference reliability characterization parameter, the candidate neural network is marked as a corresponding operation behavior analysis neural network, and the reference reliability characterization parameter can be configured according to actual needs.

[0087] Optionally, in some implementations, the step of analyzing, based on the device behavior description vector of the type and the operation behavior description vector corresponding to the smart home device to be determined, a reference device set that best matches the smart home device to be determined in the at least one reference device set, and marking the reference device set as the corresponding matching reference device set may further include the following contents:

[0088] According to the corresponding device behavior description vector of the category and the operation behavior description vector corresponding to the smart home device to be determined, the reference device set that best matches the smart home device to be determined is analyzed in the at least one reference device set through the operation behavior analysis neural network, that is, the corresponding device behavior description vector of the category and the operation behavior description vector of the smart home device to be determined are the best match;

[0089] The best matching reference device set is marked as the corresponding matching reference device set.

[0090] Optionally, in some implementations, the step of analyzing the reference device set that best matches the to-be-determined smart home device from the at least one reference device set through the operation behavior analysis neural network based on the corresponding device behavior description vector of the type and the corresponding operation behavior description vector of the to-be-determined smart home device may further include the following contents:

[0091] By using the operation behavior analysis neural network, according to the device behavior description vector of the type, analyzing the self-focus feature analysis results corresponding to the reference device set and the associated focus feature analysis results corresponding to the reference device set;

[0092] Based on the associated focus feature analysis result and the self-focus feature analysis result corresponding to the category device behavior description vector, the vector matching parameter between the category device behavior description vector and the operation behavior description vector corresponding to the to-be-determined smart home device is analyzed. Exemplarily, the vector matching parameters between the associated focus feature analysis result and the self-focus feature analysis result and the operation behavior description vector corresponding to the to-be-determined smart home device can be calculated respectively. Then, the two obtained vector matching parameters can be fused to obtain the vector matching parameter between the category device behavior description vector and the operation behavior description vector corresponding to the to-be-determined smart home device. When fusion is performed, mean calculation or weighted sum calculation can be performed;

[0093] Based on the vector matching parameters, a reference device set that best matches the smart home device to be determined is analyzed from the at least one reference device set. Exemplarily, a reference device set with the largest corresponding vector matching parameter can be used as the reference device set that best matches the smart home device to be determined.

[0094] Optionally, in some embodiments, the step of analyzing the self-focus feature analysis results and the corresponding associated focus feature analysis results corresponding to the reference device set according to the device behavior description vector by running the behavior analysis neural network may further include the following contents:

[0095] By using the operation behavior analysis neural network, according to the device behavior description vector of the type corresponding to the reference device set, a self-focus feature analysis result is analyzed, and each of the reference device sets should have a self-focus feature analysis result;

[0096] Through the operation behavior analysis neural network, based on the operation behavior description vector corresponding to the smart home device to be determined and each of the device behavior description vectors corresponding to the reference device set, the associated focus feature analysis results of each of the reference device sets are analyzed.

[0097] Optionally, in some implementations, the step of analyzing the self-focus feature analysis result by running the behavior analysis neural network according to the type device behavior description vector corresponding to the reference device set may further include the following contents:

[0098] Taking any one of the reference device sets as a first reference device set;

[0099] Based on the second parameter distribution matrix included in the operation behavior analysis neural network (the second parameter distribution matrix can be used as a network parameter of the operation behavior analysis neural network and is formed by updating), the category device behavior description vector corresponding to the first reference device set is weighted to output the corresponding weighted category behavior description vector, that is, the second parameter distribution matrix and the category device behavior description vector can be multiplied;

[0100] Performing a row-column swap operation on vector parameters of the category device behavior description vector corresponding to the first reference device set to output a corresponding swapped category behavior description vector, that is, the row vector parameters in the category device behavior description vector can be used as the column vector parameters in the swapped category behavior description vector, and the column vector parameters in the category device behavior description vector can be used as the row vector parameters in the swapped category behavior description vector;

[0101] Performing a first excitation mapping process on the weighted type behavior description vector to output a first excitation mapping description vector corresponding to the first reference device set, and performing a multiplication process on the swap type behavior description vector and the first excitation mapping description vector, and performing a second excitation mapping process on the result of the multiplication process to output a corresponding second excitation mapping description vector, wherein the first excitation mapping process is different from the second excitation mapping process, that is, the excitation function corresponding to the first excitation mapping process is different from the excitation function corresponding to the second excitation mapping process;

[0102] The category device behavior description vector corresponding to the first reference device set and the corresponding second stimulus mapping description vector are multiplied to output the self-focus feature analysis result corresponding to the first reference device set, based on which the key information of the self can be mined.

[0103] Optionally, in some implementations, the step of analyzing the associated focus feature analysis result of each reference device set through the operation behavior analysis neural network according to the operation behavior description vector corresponding to the to-be-determined smart home device and each of the device behavior description vectors of the type corresponding to the reference device set may further include the following items:

[0104] Based on the first parameter distribution included in the operation behavior analysis neural network (the first parameter distribution matrix can be used as the network parameter of the operation behavior analysis neural network and is formed by updating), the operation behavior description vector corresponding to the to-be-determined smart home device is weighted to output the corresponding weighted operation behavior description vector, that is, the first parameter distribution matrix and the operation behavior description vector can be multiplied, and, for each of the reference device sets, the row and column transposition operation of the vector parameters of the type device behavior description vector corresponding to the reference device set is performed to output the corresponding transposed device behavior description vector, that is, the row vector parameters in the type device behavior description vector can be used as the column vector parameters in the transposed device behavior description vector, and the column vector parameters in the type device behavior description vector can be used as the row vector parameters in the transposed device behavior description vector;

[0105] For each of the reference device sets, the exchange device behavior description vector corresponding to the reference device set and the weighted operation behavior description vector corresponding to the smart home device to be determined are multiplied, and the result of the multiplication is subjected to excitation mapping processing, that is, it is implemented through the corresponding excitation function to output the excitation mapping description vector corresponding to the reference device set, and the type device behavior description vector corresponding to the reference device set and the excitation mapping description vector are multiplied to output the associated focus feature analysis result corresponding to the reference device set.

[0106] Combination Figure 3 The embodiment of the present invention further provides a smart home device feature determination system, which can be applied to the smart home device feature determination platform. The smart home device feature determination system may include the following software function modules:

[0107] A description vector mining module is used to mine the operation behavior description vector corresponding to the to-be-determined smart home device, and mine the reference operation behavior description vector corresponding to the reference smart home device;

[0108] A device classification module, configured to classify a plurality of the reference smart home devices according to the reference operation behavior description vectors corresponding to the reference smart home devices, so as to form at least one corresponding reference device set, wherein any one of the at least one reference device set includes at least one reference smart home device having the same device feature type information;

[0109] A category vector determination module, configured to analyze a category device behavior description vector corresponding to each reference device set based on a reference operation behavior description vector corresponding to the reference smart home device;

[0110] a matching set determination module, configured to analyze, based on the device behavior description vector of the category and the operation behavior description vector corresponding to the smart home device to be determined, a reference device set that best matches the smart home device to be determined in the at least one reference device set, and mark the reference device set as a corresponding matching reference device set;

[0111] The device feature type determination module is used to mark the device feature type information corresponding to the matching reference device set to mark it as the device feature type information corresponding to the smart home device to be determined.

[0112] Optionally, in some implementations, the description vector mining module is specifically used to:

[0113] Collecting multi-angle device operation behavior data of the to-be-determined smart home device;

[0114] Performing a key information mining operation on the multi-angle device operation behavior data corresponding to the to-be-determined smart home device to output operation behavior description vectors of the to-be-determined smart home device at multiple angles;

[0115] The operation behavior description vectors of multiple angles corresponding to the to-be-determined smart home device are subjected to vector cascade combination processing to output the operation behavior description vector corresponding to the to-be-determined smart home device.

[0116] Optionally, in some embodiments, the device feature determination system further includes at least one other functional module, which can be used to:

[0117] Collect multi-angle device operation behavior data corresponding to the reference smart home device, and determine device feature type information corresponding to the reference smart home device;

[0118] Perform key information mining on the multi-angle device operation behavior data corresponding to the reference smart home device, and output an operation behavior description vector corresponding to the reference smart home device;

[0119] The initial neural network is updated by using typical device operation behavior data to output a corresponding candidate neural network, wherein the typical device operation behavior data includes an operation behavior description vector corresponding to the reference smart home device and corresponding device feature type information;

[0120] Performing a network reliability analysis operation on the candidate neural network to output a network reliability characterization parameter corresponding to the candidate neural network;

[0121] In the case that the network reliability characterization parameter is greater than or equal to the pre-configured reference reliability characterization parameter, the candidate neural network is marked as the corresponding operation behavior analysis neural network.

[0122] Optionally, in some implementations, the matching set determination module is specifically configured to:

[0123] According to the corresponding device behavior description vector of the type and the corresponding operation behavior description vector of the to-be-determined smart home device, a reference device set that best matches the to-be-determined smart home device is analyzed in the at least one reference device set through the operation behavior analysis neural network;

[0124] The best matching reference device set is marked as the corresponding matching reference device set.

[0125] In summary, the present invention provides a method and system for determining device features of a smart home, which can classify multiple reference smart home devices according to the reference operation behavior description vector corresponding to the reference smart home device to form at least one reference device set; analyze the type device behavior description vector corresponding to each reference device set; analyze the reference device set that best matches the smart home device to be determined according to the type device behavior description vector and the operation behavior description vector corresponding to the smart home device to be determined, and mark the reference device set as the corresponding matching reference device set; mark the device feature type information corresponding to the matching reference device set to mark it as the device feature type information corresponding to the smart home device to be determined. Based on the foregoing, on the one hand, since the matching analysis is performed on the type device behavior description vector corresponding to the reference device set, the basis is more sufficient to improve the reliability of the device feature type determination. In addition, the number of matching analyses can be reduced to make the efficiency higher.

[0126] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for determining device characteristics of a smart home, characterized in that: include: Mining out the operation behavior description vector corresponding to the to-be-determined smart home device, and mining out the reference operation behavior description vector corresponding to the reference smart home device; According to the reference operation behavior description vector corresponding to the reference smart home device, a plurality of the reference smart home devices are classified to form at least one corresponding reference device set, wherein any one of the at least one reference device set includes at least one reference smart home device having the same device feature type information; Analyze the device behavior description vector corresponding to each of the reference device sets based on the reference operation behavior description vectors corresponding to the reference smart home devices. According to the device behavior description vector of the category and the operation behavior description vector corresponding to the smart home device to be determined, in the at least one reference device set, a reference device set that best matches the smart home device to be determined is analyzed, and the reference device set is marked as a corresponding matching reference device set; Marking the device feature type information corresponding to the matching reference device set to mark it as the device feature type information corresponding to the to-be-determined smart home device; Before the step of mining the operation behavior description vector corresponding to the to-be-determined smart home device and mining the reference operation behavior description vector corresponding to the reference smart home device, the method for determining the device characteristics of the smart home further includes: Collect multi-angle device operation behavior data corresponding to the reference smart home device, and determine device feature type information corresponding to the reference smart home device; Perform key information mining on the multi-angle device operation behavior data corresponding to the reference smart home device, and output an operation behavior description vector corresponding to the reference smart home device; The initial neural network is updated by using typical device operation behavior data to output a corresponding candidate neural network, wherein the typical device operation behavior data includes an operation behavior description vector corresponding to the reference smart home device and corresponding device feature type information; Performing a network reliability analysis operation on the candidate neural network to output a network reliability characterization parameter corresponding to the candidate neural network; In the case where the network reliability characterization parameter is greater than or equal to the pre-configured reference reliability characterization parameter, marking the candidate neural network to mark it as a corresponding operation behavior analysis neural network; The step of analyzing the reference device set that best matches the to-be-determined smart home device from the at least one reference device set based on the device behavior description vector of the type and the operation behavior description vector corresponding to the to-be-determined smart home device, and marking the reference device set as the corresponding matching reference device set includes: According to the corresponding device behavior description vector of the type and the operation behavior description vector of the to-be-determined smart home device, a reference device set that best matches the to-be-determined smart home device is analyzed from the at least one reference device set through the operation behavior analysis neural network; marking the best matching reference device set as the corresponding matching reference device set; The step of analyzing the reference device set that best matches the to-be-determined smart home device from the at least one reference device set through the operation behavior analysis neural network based on the corresponding device behavior description vector of the type and the operation behavior description vector of the to-be-determined smart home device comprises: By using the operation behavior analysis neural network, according to the device behavior description vector of the type, analyzing the self-focus feature analysis results corresponding to the reference device set and the associated focus feature analysis results corresponding to the reference device set; Analyze the vector matching parameters between the device behavior description vector of the type and the operation behavior description vector corresponding to the smart home device to be determined based on the associated focus feature analysis results and the self-focus feature analysis results corresponding to the device behavior description vector of the type; Based on the vector matching parameter, analyzing the reference device set that best matches the to-be-determined smart home device from the at least one reference device set; The step of analyzing the self-focus feature analysis results and the corresponding associated focus feature analysis results corresponding to the reference device set according to the device behavior description vector by running the behavior analysis neural network includes: By using the operation behavior analysis neural network, according to the device behavior description vector of the type corresponding to the reference device set, analyzing the self-focus feature analysis result; Through the operation behavior analysis neural network, based on the operation behavior description vector corresponding to the smart home device to be determined and each of the device behavior description vectors corresponding to the reference device set, the associated focus feature analysis results of each of the reference device sets are analyzed.

2. The method for determining the characteristics of smart home devices according to claim 1, characterized in that: The step of analyzing the associated focus feature analysis result of each reference device set through the operation behavior analysis neural network according to the operation behavior description vector corresponding to the to-be-determined smart home device and each of the device behavior description vectors corresponding to the reference device set includes: Based on the first parameter distribution included in the operation behavior analysis neural network, weighted processing is performed on the operation behavior description vector corresponding to the to-be-determined smart home device to output the corresponding weighted operation behavior description vector, and, for each of the reference device sets, a row-column swap operation of vector parameters is performed on the type device behavior description vector corresponding to the reference device set to output the corresponding swapped device behavior description vector; For each of the reference device sets, the exchange device behavior description vector corresponding to the reference device set and the weighted operation behavior description vector corresponding to the smart home device to be determined are multiplied, and the result of the multiplication is subjected to excitation mapping processing to output the excitation mapping description vector corresponding to the reference device set, and the category device behavior description vector corresponding to the reference device set and the excitation mapping description vector are multiplied to output the associated focus feature analysis result corresponding to the reference device set.

3. The method for determining the characteristics of smart home devices according to claim 1, characterized in that: The step of analyzing the self-focus feature analysis result by the operation behavior analysis neural network according to the type device behavior description vector corresponding to the reference device set includes: Taking any one of the reference device sets as a first reference device set; Based on the second parameter distribution matrix included in the operation behavior analysis neural network, weighting the category device behavior description vectors corresponding to the first reference device set to output corresponding weighted category behavior description vectors; Performing a row-column swap operation on vector parameters of the category device behavior description vector corresponding to the first reference device set to output a corresponding swapped category behavior description vector; Performing a first stimulus mapping process on the weighted type behavior description vector to output a first stimulus mapping description vector corresponding to the first reference device set, and performing a multiplication process on the swap type behavior description vector and the first stimulus mapping description vector, and performing a second stimulus mapping process on the result of the multiplication process to output a corresponding second stimulus mapping description vector, wherein the first stimulus mapping process is different from the second stimulus mapping process; The category device behavior description vector corresponding to the first reference device set and the corresponding second excitation mapping description vector are multiplied to output the self-focus feature analysis result corresponding to the first reference device set.

4. The method for determining device characteristics of a smart home according to any one of claims 1 to 3, characterized in that: The step of mining the operation behavior description vector corresponding to the to-be-determined smart home device includes: Collecting multi-angle device operation behavior data of the to-be-determined smart home device; Performing a key information mining operation on the multi-angle device operation behavior data corresponding to the to-be-determined smart home device to output operation behavior description vectors of the to-be-determined smart home device at multiple angles; The operation behavior description vectors of multiple angles corresponding to the to-be-determined smart home device are subjected to vector cascade combination processing to output the operation behavior description vector corresponding to the to-be-determined smart home device.

5. A smart home device feature determination system, characterized in that: include: A description vector mining module is used to mine the operation behavior description vector corresponding to the to-be-determined smart home device, and mine the reference operation behavior description vector corresponding to the reference smart home device; A device classification module, configured to classify a plurality of the reference smart home devices according to the reference operation behavior description vectors corresponding to the reference smart home devices, so as to form at least one corresponding reference device set, wherein any one of the at least one reference device set includes at least one reference smart home device having the same device feature type information; A category vector determination module, configured to analyze a category device behavior description vector corresponding to each reference device set based on a reference operation behavior description vector corresponding to the reference smart home device; a matching set determination module, configured to analyze, based on the device behavior description vector of the category and the operation behavior description vector corresponding to the smart home device to be determined, a reference device set that best matches the smart home device to be determined in the at least one reference device set, and mark the reference device set as a corresponding matching reference device set; A device feature type determination module, used to mark the device feature type information corresponding to the matching reference device set to mark it as the device feature type information corresponding to the to-be-determined smart home device; The device feature determination system further includes at least one other functional module, which is used to: Collect multi-angle device operation behavior data corresponding to the reference smart home device, and determine device feature type information corresponding to the reference smart home device; Perform key information mining on the multi-angle device operation behavior data corresponding to the reference smart home device, and output an operation behavior description vector corresponding to the reference smart home device; The initial neural network is updated by using typical device operation behavior data to output a corresponding candidate neural network, wherein the typical device operation behavior data includes an operation behavior description vector corresponding to the reference smart home device and corresponding device feature type information; Performing a network reliability analysis operation on the candidate neural network to output a network reliability characterization parameter corresponding to the candidate neural network; In the case where the network reliability characterization parameter is greater than or equal to the pre-configured reference reliability characterization parameter, marking the candidate neural network to mark it as a corresponding operation behavior analysis neural network; The step of analyzing, based on the device behavior description vector of the type and the operation behavior description vector corresponding to the smart home device to be determined, a reference device set that best matches the smart home device to be determined in the at least one reference device set, and marking the reference device set as a corresponding matching reference device set includes: According to the corresponding device behavior description vector of the type and the operation behavior description vector of the to-be-determined smart home device, a reference device set that best matches the to-be-determined smart home device is analyzed from the at least one reference device set through the operation behavior analysis neural network; marking the best matching reference device set as the corresponding matching reference device set; Wherein, according to the corresponding device behavior description vector of the type and the operation behavior description vector of the to-be-determined smart home device, analyzing the reference device set that best matches the to-be-determined smart home device from the at least one reference device set through the operation behavior analysis neural network, comprises: By using the operation behavior analysis neural network, according to the device behavior description vector of the type, analyzing the self-focus feature analysis results corresponding to the reference device set and the associated focus feature analysis results corresponding to the reference device set; Analyze the vector matching parameters between the device behavior description vector of the type and the operation behavior description vector corresponding to the smart home device to be determined based on the associated focus feature analysis results and the self-focus feature analysis results corresponding to the device behavior description vector of the type; Based on the vector matching parameter, analyzing the reference device set that best matches the to-be-determined smart home device from the at least one reference device set; The operation behavior analysis neural network is used to analyze the self-focus feature analysis results and the corresponding associated focus feature analysis results corresponding to the reference device set according to the device behavior description vector, including: By using the operation behavior analysis neural network, according to the device behavior description vector of the type corresponding to the reference device set, analyzing the self-focus feature analysis result; Through the operation behavior analysis neural network, based on the operation behavior description vector corresponding to the smart home device to be determined and each of the device behavior description vectors corresponding to the reference device set, the associated focus feature analysis results of each of the reference device sets are analyzed.

Images