Intelligent interaction method and system based on artificial intelligence

Abstract

The invention discloses an intelligent interaction method and system based on artificial intelligence, and the method comprises the steps: building an interaction database, collecting the comprehensive data of a sensor in real time, inputting the data into the interaction database, inputting the comprehensive information of a user room into the interaction database, and collecting the historical operation data of a user. Analyzing the historical operation data to obtain an adjustment threshold value corresponding to an operation habit of the user, collecting behaviors of the user after interaction with the smart home, analyzing the behaviors to obtain behavior habit characteristics of the user, analyzing voice characteristics of the user, matching a fuzzy instruction for the voice characteristics to obtain a fuzzy instruction set, and sending the fuzzy instruction set to the smart home; the fuzzy instruction set is screened according to the adjustment threshold value, the influence of the scene feature where the user is located and the behavior habit feature on the fuzzy instructions is analyzed, the priorities of the fuzzy instructions are obtained, and the fuzzy instruction with the highest priority is selected as an interaction instruction, and the method and device have the advantages of improving accuracy and user experience.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent interaction, and particularly to an intelligent interaction method and system based on artificial intelligence. Background Art

[0002] With the rapid development of technology, various smart home devices have been popularized, making people's home life more comfortable and convenient.

[0003] Currently, most smart homes mainly rely on control terminals, voice, gestures, etc. for interaction. Among them, voice interaction is deeply favored by users for its simple and convenient operation method. However, the voice commands issued by users do not fully meet the set discrimination conditions. Existing technologies mostly use semantic judgment. However, since the voice commands issued by users may be relatively vague in semantics, such as "open the window, open the window a little, open the curtain a little, turn the air conditioner down a little", etc., existing technologies cannot accurately judge the semantics of users, resulting in inaccurate control of smart homes. Moreover, since the control of smart homes by existing technologies is often based on independent voice commands, the smart homes cannot be controlled associatively, and it is easy for users to need to issue commands frequently, resulting in a poor interaction experience of smart homes. Therefore, it is necessary to design an intelligent interaction method and system based on artificial intelligence to improve accuracy and user experience. Summary of the Invention

[0004] The purpose of the present invention is to provide an intelligent interaction method and system based on artificial intelligence to solve the problems raised in the above background art.

[0005] To solve the above technical problems, the present invention provides the following technical solutions: An intelligent interaction method based on artificial intelligence, including:

[0006] Establish an interaction database, collect sensor comprehensive data in real time and enter the data into the interaction database, and enter the comprehensive information of the user's room into the interaction database;

[0007] Analyze the historical operation data of the user, further analyze the adjustment threshold corresponding to the user's operation habits to obtain the user's habitual adjustment threshold, analyze the user's behavior characteristics to obtain the user's behavior habit characteristics;

[0008] Analyze the voice characteristics of the user, match the fuzzy commands, further analyze the scene where the user is located, analyze the scene environment characteristics, analyze the influence of the scene environment characteristics on the fuzzy commands, evaluate the priority of the fuzzy commands, and determine the interaction commands;

[0009] Analyze the changes in the user's behavior and scene environment, analyze the linkage commands of the interaction commands, match the active commands, and interact with the smart home according to the commands.

[0010] According to the above technical solution, analyzing the historical operation data of the user, further analyzing the adjustment threshold corresponding to the user's operation habit to obtain the user's habit adjustment threshold, and analyzing the user's behavior characteristics to obtain the user's behavior habit characteristics include the following steps:

[0011] Obtain the user's historical operation data, identify the timestamps of the user's historical voice commands, identify the target smart home of the historical voice commands with adjacent timestamps. If the target smart homes are the same, sort the historical voice commands in descending order according to the timestamps, anchor the last historical voice command, retrieve the corresponding adjustment threshold in the database according to the last historical voice command, bind the adjustment threshold corresponding to the first historical voice command and the last historical voice command, identify the characteristics of the first historical voice command, and perform union fusion on the adjustment thresholds with the same characteristics to obtain the user's habit adjustment threshold;

[0012] Anchor the timestamps of the first historical voice command and the last historical voice command, identify the time period T0 between the first historical voice command and the last historical voice command, retrieve the corresponding visual image in the database according to the time period T0, identify the user's edge nodes, construct a user body model, identify the action characteristics of the body model, construct a coordinate system, mark the user's position and the smart home's position in the coordinate system, record the relative position between the user and the smart home, identify the relative position characteristics, fuse the action characteristics and the relative position characteristics to obtain the user's historical behavior characteristics, obtain the user's historical behavior characteristics corresponding to each historical cloud factor command in the historical operation data, identify the frequency of occurrence of the user's historical behavior characteristics, retrieve the set weighted fusion ratio in the database according to the frequency, and perform weighted fusion on the user's historical behavior characteristics according to the ratio to obtain the user's behavior habit characteristics.

[0013] According to the above technical solution, analyzing the user's voice characteristics, matching fuzzy commands, further analyzing the user's location scenario, analyzing the scenario environment characteristics, analyzing the influence of the scenario environment characteristics on the fuzzy commands, and evaluating the priority of the fuzzy commands to determine the interaction commands include the following steps:

[0014] Obtain the user's voice command, identify the semantics of the user's voice command, retrieve the set command set in the database and compare it. If the similarity is greater than the threshold, mark the first command with the greatest similarity as the interaction command. Otherwise, extract the keywords in the user's voice command, retrieve the corresponding fuzzy commands in the database according to the keywords, and obtain a set of fuzzy commands;

[0015] Obtain comprehensive room information, identify the three-dimensional data of the room, construct a three-dimensional model of the room, identify the timestamp of the user's voice command, retrieve the database according to the timestamp of the user's voice command, retrieve the corresponding comprehensive room model, obtain the visual image in the room, anchor the curtain image in the visual image, identify the curtain edge nodes in the curtain image, and mark the curtain edge nodes in the three-dimensional room model;

[0016] Identify the room light irradiation data, construct virtual irradiation points according to the room light irradiation data, fit the three-dimensional room model and the virtual irradiation points to obtain a comprehensive room model;

[0017] Establish a three-dimensional coordinate system, identify the coordinate point a of the virtual irradiation point and the coordinate point b of the curtain edge node in the comprehensive room model, connect the coordinate point a and the coordinate point b to construct a light irradiation model, identify the slope of the connection line ab, if the slope is greater than the set threshold, then identify the fuzzy command to pull the right curtain in the fuzzy command set and mark it as the right fuzzy command, otherwise identify the fuzzy command to pull the left curtain in the fuzzy command set and mark it as the left fuzzy command.

[0018] According to the above technical solution, the steps of analyzing the user's voice characteristics, matching fuzzy commands, further analyzing the user's location scenario, analyzing the scene environment characteristics, analyzing the influence of the scene environment characteristics on the fuzzy commands, evaluating the priority of the fuzzy commands, and determining the interaction commands further include the following steps:

[0019] Obtain the fuzzy command set, perform simulated control on the curtain in the comprehensive room model according to the fuzzy command set, and output the room model A after adjusting the curtain according to the fuzzy command. Identify the edge inflection points c, d, e, f of the curtain, where the edge inflection points c, d respectively represent the upper and lower edge inflection points inside the left curtain, and the edge inflection points e, f respectively represent the upper and lower edge inflection points inside the right curtain. Mark the edge inflection points of the curtain in the room model A, and connect the virtual irradiation points with the edge inflection points c, d, e, f respectively to construct the coordinate connection lines between the virtual irradiation points and the edge inflection points;

[0020] Obtain the visual image in the room, identify the user contour nodes, construct a user model, identify the plane where the user model is located, identify the intersection point of the connection line between the virtual irradiation point and the edge inflection point coordinates and the plane where the user is located, connect adjacent intersection points, fit the connection lines of adjacent intersection points, construct a light radiation model, identify the included angle θ1 between adjacent sides of the light radiation model, identify the coordinates of the adjacent intersection points, calculate the lengths l1 and l2 of the connection lines between the adjacent intersection points after fitting through the distance formula, where l1 and l2 represent two adjacent sides in the light radiation model, calculate the area S1 of the light radiation model through the formula S1 = α × sinθ1 × l1 × l2, where α represents the error coefficient of the area of the light radiation model, and retrieve the corresponding influence coefficient β of the fuzzy instruction from the database according to the area of the light radiation model i , where i = 1, 2, 3......n, β i represents the influence coefficient of the i-th fuzzy instruction on the area of the light radiation model after curtain regulation on the fuzzy instruction;

[0021] Identify the body area where the user is used to sunbathe, anchor the area of the light radiation model, perform an overlapping comparison on the areas, mark the user body area in the overlapping area as the first area, identify the area S2 of the first area, and calculate the ratio of the area of the first area to the area of the light radiation model Retrieve the corresponding influence coefficient δ from the database according to the ratio;

[0022] Obtain the user's behavior habit characteristics, identify the user's behavior habits. If the user is used to sunbathe, retrieve the influence weight ε1 of the set influence coefficient δ from the database, otherwise retrieve the influence weight ε2 of the set influence coefficient δ from the database;

[0023] Obtain the fuzzy instruction, identify the curtain adjustment distance l3 corresponding to the fuzzy instruction, compare it with the user's habitual adjustment threshold. If the curtain adjustment distance l3 is greater than the user's habitual adjustment threshold, delete the fuzzy instruction, otherwise the system continues to detect;

[0024] Obtain the remaining fuzzy instructions, obtain the basic priority score P0 set by the fuzzy instructions, calculate the priority score P1 of the remaining fuzzy instructions through the formula P1 = β × δ × P0, sort the priority scores of the remaining fuzzy instructions in descending order, select the first-ranked fuzzy instruction, and generate an interaction instruction according to the first-ranked fuzzy instruction.

[0025] According to the above technical solution, analyzing the user's behavior and changes in the scene environment, analyzing the linkage instructions of the interaction instructions, matching the active instructions, and interacting with the smart home according to the instructions include the following steps:

[0026] Obtain an interaction instruction, identify the markers in the interaction instruction. If the interaction instruction is a left-side blur instruction, open the left-side curtain according to the interaction instruction; otherwise, open the right-side curtain.

[0027] Obtain the real-time time point, retrieve the corresponding room comprehensive model B in the database according to the time point, identify the coordinates of the virtual irradiation point in the room comprehensive model B, identify the coordinates of the first curtain edge node, connect the virtual irradiation point and the first curtain edge node, identify the first slope of the connection line. If the first slope is greater than the threshold, obtain the right-side blur instruction, identify the coordinates of the second curtain edge node after adjustment of the right-side blur instruction, connect the virtual irradiation point and the second curtain edge node, identify the second slope of the connection line, compare with the database, anchor the right-side blur instruction corresponding to the second slope that meets the threshold, and use the right-side blur instruction as the first active instruction; otherwise, obtain the left-side blur instruction.

[0028] According to the above technical solution, the intelligent interaction system based on artificial intelligence includes: a data acquisition module, a behavior recognition module, an instruction analysis module, and an interaction control module;

[0029] The data acquisition module is used to establish an interaction database, collect sensor comprehensive data in real time and enter the data into the interaction database, and enter the user room comprehensive information into the interaction database;

[0030] The behavior recognition module is used to analyze the historical operation data of the user, further analyze the adjustment threshold corresponding to the user's operation habits to obtain the user's habit adjustment threshold, and analyze the user's behavior characteristics to obtain the user's behavior habit characteristics;

[0031] The instruction analysis module is used to analyze the voice characteristics of the user, match the blur instruction, further analyze the scene where the user is located, analyze the scene environment characteristics, analyze the influence of the scene environment characteristics on the blur instruction, evaluate the priority of the blur instruction, and determine the interaction instruction;

[0032] The interaction control module is used to analyze the changes in the user's behavior and the scene environment, analyze the linkage instructions of the interaction instruction, match the active instruction, and interact with the smart home according to the instruction.

[0033] According to the above technical solution, the data acquisition module includes a sensor module and a room comprehensive information entry module;

[0034] The sensor module is used to establish an interaction database, collect sensor comprehensive data in real time and enter the data into the interaction database;

[0035] The room comprehensive information entry module is used to enter the user room comprehensive information into the interaction database.

[0036] According to the above technical solution, the behavior recognition module includes a habit analysis module and a behavior analysis module;

[0037] The habit analysis module is used to obtain the user's historical operation data, identify the timestamps of the user's historical voice commands, identify the target smart home of the historical voice commands with adjacent timestamps. If the target smart homes are the same, the historical voice commands are sorted in descending order according to the timestamps, the last historical voice command is anchored, the corresponding adjustment threshold is retrieved from the database according to the last historical voice command, the adjustment threshold corresponding to the first historical voice command is bound to the adjustment threshold corresponding to the last historical voice command, the features of the first historical voice command are identified, and the adjustment thresholds with the same features are merged by union to obtain the user habit adjustment threshold;

[0038] The behavior analysis module is used to anchor the timestamps of the first historical voice command and the last historical voice command, identify the time period T0 between the first historical voice command and the last historical voice command, retrieve the corresponding visual image from the database according to the time period T0, identify the user's edge nodes, construct a user body model, identify the action features of the body model, construct a coordinate system, mark the user's position and the smart home's position in the coordinate system, record the relative position between the user and the smart home, identify the relative position features, and merge the action features with the relative position features to obtain the user's historical behavior features. Obtain the user's historical behavior features corresponding to each historical cloud command in the historical operation data, identify the frequency of occurrence of the user's historical behavior features, retrieve the set weighted fusion ratio from the database according to the frequency, and perform weighted fusion on the user's historical behavior features according to the ratio to obtain the user's behavior habit features.

[0039] According to the above technical solution, the instruction analysis module includes a fuzzy instruction matching module, a scenario analysis module, and a priority evaluation module;

[0040] The fuzzy instruction matching module is used to obtain the user's voice command, identify the semantics of the user's voice command, retrieve the set instruction set in the data and compare it. If the similarity is greater than the threshold, the first instruction with the greatest similarity is marked as the interaction instruction. Otherwise, the keywords in the user's voice command are extracted, the database is retrieved according to the keywords, and the corresponding fuzzy instruction is retrieved from the database to obtain a fuzzy instruction set;

[0041] The scenario analysis module is used to obtain comprehensive room information, identify three-dimensional room data, construct a three-dimensional room model, identify the timestamp of the user's voice command, retrieve the database according to the timestamp of the user's voice command, retrieve the corresponding comprehensive room model, obtain the visual image in the room, anchor the curtain image in the visual image, identify the curtain edge nodes in the curtain image, and mark the curtain edge nodes in the three-dimensional room model;

[0042] Identify the room light irradiation data, construct virtual irradiation points according to the room light irradiation data, and fit the three-dimensional room model and the virtual irradiation points to obtain a comprehensive room model;

[0043] Establish a three-dimensional coordinate system, identify the coordinate point a of the virtual irradiation point and the coordinate point b of the curtain edge node in the comprehensive room model, connect the coordinate point a and the coordinate point b to construct a light irradiation model, identify the slope of the connection line ab, if the slope is greater than the set threshold, then identify the fuzzy command to pull the right curtain in the fuzzy command set and mark it as the right fuzzy command, otherwise identify the fuzzy command to pull the left curtain in the fuzzy command set and mark it as the left fuzzy command;

[0044] The priority evaluation module is used to obtain the fuzzy command set, simulate and control the curtain in the comprehensive room model according to the fuzzy command set, and output the room model A after adjusting the curtain according to the fuzzy command. Identify the edge inflection points c, d, e, f of the curtain, where the edge inflection points c, d respectively represent the upper and lower edge inflection points inside the left curtain, and the edge inflection points e, f respectively represent the upper and lower edge inflection points inside the right curtain. Mark the edge inflection points of the curtain in the room model A, connect the virtual irradiation points with the edge inflection points c, d, e, f respectively, and construct the coordinate connection lines between the virtual irradiation points and the edge inflection points;

[0045] Obtain the visual image in the room, identify the user contour nodes, construct a user model, identify the plane where the user model is located, identify the intersection points of the coordinate connection lines between the virtual irradiation points and the edge inflection points and the plane where the user is located, connect the adjacent intersection points, fit the adjacent intersection point connection lines, construct a light radiation model, identify the included angle θ1 between the adjacent sides of the light radiation model, identify the coordinates of the adjacent intersection points, calculate the lengths l1 and l2 of the connection lines between the adjacent intersection points after fitting through the distance formula, where l1 and l2 represent two adjacent sides in the light radiation model, calculate the area S1 of the light radiation model through the formula S1 = α×sinθ1×l1×l2, where α represents the error coefficient of the area of the light radiation model, and retrieve the corresponding influence coefficient β of the fuzzy command from the database according to the area of the light radiation model i where, i = 1, 2, 3......n, β iIt represents the influence coefficient of the area of the light radiation model after the i-th fuzzy instruction adjusts the curtain on the fuzzy instruction.

[0046] Identify the body area where the user is used to sunbathe, anchor the area of the light radiation model, overlap and compare the areas, mark the user's body area in the overlapping area as the first area, identify the area S2 of the first area, and calculate the ratio of the area of the first area to the area of the light radiation model. Retrieve the database according to the ratio and retrieve the corresponding influence coefficient δ of the fuzzy instruction.

[0047] Obtain the user's behavior habit characteristics, identify the user's behavior habits. If the user is used to sunbathe, retrieve the influence weight ε1 of the set influence coefficient δ in the database, otherwise retrieve the influence weight ε2 of the set influence coefficient δ in the database.

[0048] Obtain the fuzzy instruction, identify the curtain adjustment distance l3 corresponding to the fuzzy instruction, compare it with the user's habitual adjustment threshold. If the curtain adjustment distance l3 is greater than the user's habitual adjustment threshold, delete the fuzzy instruction, otherwise the system continues to detect.

[0049] Obtain the remaining fuzzy instructions, obtain the basic priority score P0 set by the fuzzy instructions, calculate the priority score P1 of the remaining fuzzy instructions through the formula P1 = β×δ×P0, sort the priority scores of the remaining fuzzy instructions in descending order, select the first-ranked fuzzy instruction, and generate an interaction instruction according to the first-ranked fuzzy instruction.

[0050] According to the above technical solution, the interaction control module is also used to obtain the interaction instruction, identify the mark in the interaction instruction. If the interaction instruction is a left-side fuzzy instruction, open the left-side curtain according to the interaction instruction, otherwise open the right-side curtain.

[0051] Obtain the real-time time point, retrieve the corresponding room comprehensive model B in the database according to the time point, identify the coordinates of the virtual irradiation point in the room comprehensive model B, identify the coordinates of the first curtain edge node, connect the virtual irradiation point and the first curtain edge node, identify the first slope of the connection line. If the first slope is greater than the threshold, obtain the right-side fuzzy instruction, identify the coordinates of the second curtain edge node after the right-side fuzzy instruction is adjusted, connect the virtual irradiation point and the second curtain edge node, identify the second slope of the connection line, compare with the database, anchor the right-side fuzzy instruction corresponding to the second slope that meets the threshold, and use the right-side fuzzy instruction as the first active instruction, otherwise obtain the left-side fuzzy instruction.

[0052] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: In the present invention, by analyzing the user's historical operation records, the adjustment threshold for the smart home is analyzed, and further, the adjustment threshold corresponding to the user's fuzzy instruction can be accurately determined, avoiding the phenomenon of large adjustment fluctuations. At the same time, the user's behavior during the adjustment of the smart home is analyzed to obtain the user's behavior habit characteristics, so that subsequent adjustments to the smart home can more accurately conform to the user's habits and meet the user's personalized needs. By analyzing the slope of the light model, the light irradiation direction is judged, and further, the side where the curtain is to be opened is judged, thus avoiding the wrong judgment of the curtain opening direction and affecting the user experience. The intention expressed by the user's voice instruction can be accurately judged, thereby greatly improving the accuracy of voice instruction judgment. By analyzing the projected area of the light model passing through the gap where the curtain is opened in the room, and further analyzing the ratio of the area of the overlapping part between the user's body and the light radiation model to the area of the light radiation model, the impact of the light radiating through the curtain in the room on the user experience can be accurately analyzed. Further analyzing the user's behavior habit characteristics, determining the user's habits, and adjusting the weight of the influence coefficient, the priority of the fuzzy instruction can be accurately evaluated, thereby greatly improving the accuracy of the system and avoiding affecting the user experience and improving the user's comfort. By using the above method to determine the second active instruction, the curtain is controlled according to the active instruction. By analyzing the change of the virtual irradiation point, further analyzing the slope of the line connecting the virtual irradiation point and the curtain edge node, and further determining the fuzzy instruction corresponding to the qualified slope, an active instruction is generated to control the curtain, which can avoid the user from frequently issuing instructions to control the curtain, thereby greatly improving the user's experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] The drawings are used to provide a further understanding of the present invention and constitute a part of the specification. They are used together with the embodiments of the present invention to explain the present invention and do not constitute a limitation to the present invention. In the drawings:

[0054] Figure 1 is a flowchart of the method steps of the present invention.

[0055] Figure 2 is a schematic diagram of the system module composition of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0056] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0057] Please refer to Figure 1, the present invention provides a technical solution: an intelligent interaction method based on artificial intelligence, including:

[0058] Step S1: Establish an interaction database, collect comprehensive sensor data in real time and input the data into the interaction database, and input the comprehensive information of the user's room into the interaction database;

[0059] Step S2: Analyze the user's historical operation data, further analyze the adjustment threshold corresponding to the user's operation habits to obtain the user's habit adjustment threshold, and analyze the user's behavior characteristics to obtain the user's behavior habit characteristics;

[0060] Step S3: Analyze the user's voice characteristics, match fuzzy instructions for the voice characteristics to obtain a set of fuzzy instructions, screen the set of fuzzy instructions according to the adjustment threshold, analyze the influence of the user's current scene characteristics and the behavior habit characteristics on the fuzzy instructions and obtain the priority of the fuzzy instructions, and select the fuzzy instruction with the highest priority as the interaction instruction;

[0061] Step S4: Analyze the changes in the user's behavior and the scene environment, analyze the linkage instructions of the interaction instructions, match the active instructions, and interact with the smart home according to the instructions.

[0062] In the present invention, by analyzing the slope of the light model, the light irradiation direction is judged, and further the side where the curtain is opened is judged, so as to avoid misjudgment of the curtain opening direction and affect the user experience. The intention expressed by the user's voice command can be accurately judged, and thus the accuracy of voice command judgment is greatly improved. By analyzing the projected area of the light model passing through the gap opened by the curtain in the room, and further analyzing the ratio of the overlapping area between the user's body and the light radiation model to the area of the light radiation model, the influence of the light radiating through the curtain in the room on the user experience can be accurately analyzed. By further analyzing the user's behavior habit characteristics, determining the user's habits, and adjusting the weight of the influence coefficient, the priority of the fuzzy instructions can be accurately evaluated, and thus the accuracy of the system is greatly improved, while avoiding affecting the user experience and improving the user's comfort.

[0063] In some preferred embodiments, step S2 further includes the following steps:

[0064] Step S21: Obtain the user's historical operation data, identify the timestamps of the user's historical voice commands, identify the target smart home of the historical voice commands with adjacent timestamps. If the target smart homes are the same, sort the historical voice commands in descending order according to the timestamps, anchor the last historical voice command, retrieve the corresponding adjustment threshold from the database according to the last historical voice command, bind the adjustment threshold corresponding to the first historical voice command to the adjustment threshold corresponding to the last historical voice command, identify the features of the first historical voice command, and perform union fusion on the adjustment thresholds with the same features to obtain the user's habit adjustment threshold;

[0065] Step S22: Anchor the timestamps of the first historical voice command and the last historical voice command, identify the time period T0 between the first historical voice command and the last historical voice command, retrieve the corresponding visual image from the database according to the time period T0, identify the user's edge nodes, construct a user body model, identify the action features of the body model, construct a coordinate system, mark the user's position and the smart home's position in the coordinate system, record the relative position between the user and the smart home, identify the relative position features, and fuse the action features and the relative position features to obtain the user's historical behavior features. Obtain the user's historical behavior features corresponding to each historical cloud factor command in the historical operation data, identify the frequency of occurrence of the user's historical behavior features, retrieve the set weighted fusion ratio from the database according to the frequency, and perform weighted fusion on the user's historical behavior features according to the ratio to obtain the user's behavior habit features. By analyzing the user's historical operation records and analyzing the adjustment threshold of the smart home by the user, it is further possible to accurately determine the adjustment threshold corresponding to the user's fuzzy command, avoid the phenomenon of large adjustment fluctuations, and at the same time analyze the user's behavior during the adjustment of the smart home to obtain the user's behavior habit features, so that the subsequent adjustment of the smart home can more accurately conform to the user's habits and meet the user's personalized needs.

[0066] In some preferred embodiments, step S3 further includes the following steps:

[0067] Step S31: Obtain the user's voice command, identify the semantics of the user's voice command, retrieve the set instruction set in the data and compare it. If the similarity is greater than the threshold, mark the first command with the greatest similarity as the interaction command. Otherwise, extract the keywords in the user's voice command, retrieve the corresponding fuzzy command from the database according to the keywords, and obtain the fuzzy command set;

[0068] Step S32: Obtain the comprehensive room information, identify the three-dimensional data of the room, construct a three-dimensional model of the room, identify the timestamp of the user's voice command, retrieve the database according to the timestamp of the user's voice command, retrieve the corresponding comprehensive room model, obtain the visual image in the room, anchor the curtain image in the visual image, identify the curtain edge nodes in the curtain image, and mark the curtain edge nodes in the three-dimensional model of the room;

[0069] Identify the room light irradiation data, construct virtual irradiation points according to the room light irradiation data, and fit the three-dimensional model of the room and the virtual irradiation points to obtain a comprehensive room model;

[0070] Step S33: Establish a three-dimensional coordinate system, identify the coordinate point a of the virtual irradiation point and the coordinate point b of the curtain edge node in the comprehensive room model, connect the coordinate point a and the coordinate point b to construct a light irradiation model, identify the slope of the connection line ab. If the slope is greater than the set threshold, it means that the right side of the curtain is pulled up, identify the fuzzy command to pull up the right side curtain in the fuzzy command set and mark it as the right-side fuzzy command. Otherwise, identify the fuzzy command to pull up the left side curtain in the fuzzy command set and mark it as the left-side fuzzy command. By analyzing the slope of the light model, judge the light irradiation direction, and further judge the side where the curtain is opened, so as to avoid misjudging the curtain opening direction and affecting the user experience, and be able to accurately judge the intention expressed by the user's voice command, thereby greatly improving the accuracy of voice command judgment.

[0071] In some preferred embodiments, step S33 further includes the following steps:

[0072] Step S331: Obtain the fuzzy command set, perform simulation control on the curtain in the comprehensive room model according to the fuzzy command set, and output the room model A after adjusting the curtain according to the fuzzy command. Identify the edge inflection points c, d, e, f of the curtain, where the edge inflection points c, d respectively represent the upper and lower edge inflection points inside the left curtain, and the edge inflection points e, f respectively represent the upper and lower edge inflection points inside the right curtain. Mark the edge inflection points of the curtain in the room model A, and connect the virtual irradiation points with the edge inflection points c, d, e, f respectively to construct the coordinate connection lines between the virtual irradiation points and the edge inflection points;

[0073] Step S332: Obtain the visual image in the room, identify the user contour nodes, construct a user model, identify the plane where the user model is located, identify the intersection point of the connection line between the virtual irradiation point and the edge inflection point coordinates and the plane where the user is located, connect adjacent intersection points, fit the connection lines of adjacent intersection points, construct a light radiation model, identify the included angle θ1 between adjacent sides of the light radiation model, identify the coordinates of the adjacent intersection points, calculate the lengths l1 and l2 of the connection lines between the adjacent intersection points after fitting through the distance formula, where l1 and l2 represent two adjacent sides in the light radiation model, calculate the area S1 of the light radiation model through the formula S1 = α×sinθ1×l1×l2, where α represents the error coefficient of the area of the light radiation model, and retrieve the corresponding influence coefficient β of the fuzzy instruction from the database according to the area of the light radiation model i , where i = 1, 2, 3......n, β i represents the influence coefficient of the area of the light radiation model after the i-th fuzzy instruction adjusts the curtain on the fuzzy instruction;

[0074] Step S333: Identify the body area where the user is used to sunbathe, anchor the area of the light radiation model, perform an overlapping comparison on the areas, mark the user body area in the overlapping area as the first area, identify the area S2 of the first area, and calculate the ratio of the area of the first area to the area of the light radiation model Retrieve the corresponding influence coefficient δ from the database according to the ratio;

[0075] Obtain the user's behavior habit characteristics, identify the user's behavior habits. If the user is used to sunbathe, retrieve the influence weight ε1 of the set influence coefficient δ from the database, otherwise retrieve the influence weight ε2 of the set influence coefficient δ from the database. By analyzing the projected area of the light model shining into the room through the gap when the curtain is opened, and further analyzing the ratio of the area where the user's body coincides with the light radiation model to the area of the light radiation model, it is possible to accurately analyze the impact of the light radiating through the curtain in the room on the user experience, further analyze the user's behavior habit characteristics, determine the user's habits, adjust the weight of the influence coefficient, accurately evaluate the priority of the fuzzy instruction, thereby greatly improving the accuracy of the system, avoiding affecting the user experience at the same time, and improving the user's comfort;

[0076] Obtain the fuzzy instruction, identify the curtain adjustment distance l3 corresponding to the fuzzy instruction, compare it with the user's habitual adjustment threshold. If the curtain adjustment distance l3 is greater than the user's habitual adjustment threshold, delete the fuzzy instruction, otherwise the system continues to detect;

[0077] Obtain the remaining fuzzy instructions, obtain the basic priority score P0 set by the fuzzy instructions, calculate the priority score P1 of the remaining fuzzy instructions through the formula P1 = β×δ×P0, sort the priority scores of the remaining fuzzy instructions in descending order, select the fuzzy instruction in the first place, and generate an interaction instruction according to the fuzzy instruction in the first place.

[0078] In some preferred embodiments, step S4 further includes the following steps:

[0079] Step S41: Obtain the interaction instruction, identify the mark in the interaction instruction. If the interaction instruction is a left fuzzy instruction, open the left curtain according to the interaction instruction; otherwise, open the right curtain.

[0080] Step S42: Obtain the real-time time point, retrieve the corresponding room comprehensive model B in the database according to the time point, identify the coordinates of the virtual irradiation point in the room comprehensive model B, identify the coordinates of the first curtain edge node, connect the virtual irradiation point and the first curtain edge node, identify the first slope of the connection line. If the first slope is greater than the threshold, obtain the right fuzzy instruction, identify the coordinates of the second curtain edge node after adjustment of the right fuzzy instruction, connect the virtual irradiation point and the second curtain edge node, identify the second slope of the connection line, compare with the database, anchor the right fuzzy instruction corresponding to the second slope that meets the threshold, and use the right fuzzy instruction as the first active instruction; otherwise, obtain the left fuzzy instruction, determine the second active instruction through the above method, control the curtain according to the active instruction, and by analyzing the change of the virtual irradiation point, further analyze the slope of the connection line between the virtual irradiation point and the curtain edge node, and further determine the fuzzy instruction corresponding to the slope that meets the conditions, generate the active instruction to control the curtain, which can avoid the user needing to frequently issue instructions to control the curtain, thereby greatly improving the user experience.

[0081] With the same inventive concept as the above embodiment, the present application also provides an intelligent interaction system based on artificial intelligence, including: a data acquisition module, a behavior recognition module, an instruction analysis module, and an interaction control module;

[0082] The data acquisition module is used to establish an interaction database, collect sensor comprehensive data in real time and enter the data into the interaction database, and enter the user room comprehensive information into the interaction database;

[0083] The behavior recognition module is used to analyze the historical operation data of the user, further analyze the adjustment threshold corresponding to the user's operation habit to obtain the user's habit adjustment threshold, and analyze the user's behavior characteristics to obtain the user's behavior habit characteristics;

[0084] The instruction analysis module is used to analyze the user's voice features, match fuzzy instructions for the voice features, obtain a set of fuzzy instructions, screen the set of fuzzy instructions according to the adjustment threshold, analyze the influence of the user's scene features and the behavior habit features on the fuzzy instructions and obtain the priority of the fuzzy instructions, and select the fuzzy instruction with the highest priority as the interaction instruction;

[0085] The interaction control module is used to analyze the changes in the user's behavior and the scene environment, analyze the linkage instructions of the interaction instructions, match the active instructions, and interact with the smart home according to the instructions.

[0086] The data acquisition module includes a sensor module and a room comprehensive information input module;

[0087] The sensor module is used to establish an interaction database, collect sensor comprehensive data in real time and input the data into the interaction database;

[0088] The room comprehensive information input module is used to input the user's room comprehensive information into the interaction database.

[0089] The behavior recognition module includes a habit analysis module and a behavior analysis module;

[0090] The habit analysis module is used to obtain the user's historical operation data, identify the time stamps of the user's historical voice instructions, identify the target smart home of the historical voice instructions with adjacent time stamps. If the target smart homes are the same, sort the historical voice instructions in descending order according to the time stamps, anchor the last historical voice instruction, retrieve the database according to the last historical voice instruction, retrieve the corresponding adjustment threshold in the database, bind the adjustment threshold corresponding to the first historical voice instruction to the adjustment threshold corresponding to the last historical voice instruction, identify the features of the first historical voice instruction, and perform union fusion on the adjustment thresholds with the same features to obtain the user habit adjustment threshold;

[0091] The behavior analysis module is used to anchor the timestamps of the historical voice commands in the first order and the historical voice commands in the last order, identify the time period T0 between the historical voice commands in the first order and the historical voice commands in the last order, retrieve the corresponding visual images in the database according to the time period T0, identify the edge nodes of the user, construct a user body model, identify the action features of the body model, construct a coordinate system, mark the user's position and the smart home's position in the coordinate system, record the relative position between the user and the smart home, identify the relative position features, fuse the action features and the relative position features to obtain the user's historical behavior features, obtain the user's historical behavior features corresponding to each historical cloud factor command in the historical operation data, identify the frequency of occurrence of the user's historical behavior features, retrieve the set weighted fusion ratio in the database according to the frequency, and perform weighted fusion on the user's historical behavior features according to the ratio to obtain the user's behavior habit features.

[0092] The instruction analysis module includes a fuzzy instruction matching module, a scenario analysis module, and a priority evaluation module;

[0093] The fuzzy instruction matching module is used to obtain the user's voice instruction, identify the semantics of the user's voice instruction, retrieve the set instruction set in the data and compare them. If the similarity is greater than the threshold, mark the first instruction with the maximum similarity as the interaction instruction. Otherwise, extract the keywords in the user's voice instruction, retrieve the database according to the keywords, and retrieve the corresponding fuzzy instruction in the database to obtain a fuzzy instruction set;

[0094] The scenario analysis module is used to obtain the comprehensive room information, identify the three-dimensional room data, construct a three-dimensional room model, identify the timestamp of the user's voice instruction, retrieve the corresponding comprehensive room model in the database according to the timestamp of the user's voice instruction, obtain the visual image in the room, anchor the curtain image in the visual image, identify the curtain edge nodes in the curtain image, and mark the curtain edge nodes in the three-dimensional room model;

[0095] Identify the room light irradiation data, construct a virtual irradiation point according to the room light irradiation data, and fit the three-dimensional room model and the virtual irradiation point to obtain a comprehensive room model;

[0096] Establish a three-dimensional coordinate system, identify the coordinate point a of the virtual irradiation point and the coordinate point b of the curtain edge node in the comprehensive room model, connect the coordinate point a and the coordinate point b to construct a light irradiation model, identify the slope of the connection line ab. If the slope is greater than the set threshold, it means that the right side of the curtain is pulled up. Identify the fuzzy instruction to pull up the right side curtain in the fuzzy instruction set and mark it as the right-side fuzzy instruction. Otherwise, identify the fuzzy instruction to pull up the left side curtain in the fuzzy instruction set and mark it as the left-side fuzzy instruction;

[0097] The priority evaluation module is used to obtain a set of fuzzy instructions, simulate the control of the curtain in the room comprehensive model according to the set of fuzzy instructions, and output the room model A after adjusting the curtain according to the fuzzy instructions. Identify the edge inflection points c, d, e, and f of the curtain, where the edge inflection points c and d respectively represent the upper and lower edge inflection points inside the left curtain, and the edge inflection points e and f respectively represent the upper and lower edge inflection points inside the right curtain. Mark the edge inflection points of the curtain in the room model A, connect the virtual irradiation points with the edge inflection points c, d, e, and f respectively, and construct the coordinate connection lines between the virtual irradiation points and the edge inflection points.

[0098] Obtain the visual image in the room, identify the user contour nodes, construct the user model, identify the plane where the user model is located, identify the intersection points of the coordinate connection lines between the virtual irradiation points and the edge inflection points and the plane where the user is located, connect the adjacent intersection points, fit the adjacent intersection point connection lines, construct the light radiation model, identify the included angle θ1 between the adjacent sides of the light radiation model, identify the coordinates of the adjacent intersection points, calculate the lengths l1 and l2 of the connection lines between the adjacent intersection points after fitting through the distance formula, where l1 and l2 represent two adjacent sides in the light radiation model, and calculate the area S1 of the light radiation model through the formula S1 = α × sinθ1 × l1 × l2, where α represents the error coefficient of the area of the light radiation model. Retrieve the corresponding influence coefficient β of the fuzzy instruction from the database according to the area of the light radiation model. i , where i = 1, 2, 3......n, β i represents the influence coefficient of the area of the light radiation model after the i-th fuzzy instruction adjusts the curtain on the fuzzy instruction.

[0099] Identify the body area where the user is used to sunbathe, anchor the area of the light radiation model, perform an overlapping comparison on the areas, mark the user body area in the overlapping area as the first area, identify the area S2 of the first area, and calculate the ratio of the area of the first area to the area of the light radiation model. Retrieve the corresponding influence coefficient δ from the database according to the ratio.

[0100] Obtain the user's behavior habit characteristics, identify the user's behavior habits. If the user is used to sunbathe, retrieve the influence weight ε1 of the set influence coefficient δ from the database, otherwise retrieve the influence weight ε2 of the set influence coefficient δ from the database.

[0101] Obtain the fuzzy instruction, identify the curtain adjustment distance l3 corresponding to the fuzzy instruction, compare it with the user's habitual adjustment threshold. If the curtain adjustment distance l3 is greater than the user's habitual adjustment threshold, delete the fuzzy instruction, otherwise the system continues to detect.

[0102] Obtain the remaining fuzzy instructions, obtain the basic priority score P0 set by the fuzzy instructions, calculate the priority score P1 of the remaining fuzzy instructions through the formula P1 = β×δ×P0, sort the priority scores of the remaining fuzzy instructions in descending order, select the fuzzy instruction in the first place, and generate an interaction instruction according to the fuzzy instruction in the first place.

[0103] The interaction control module is further configured to obtain an interaction instruction, identify the mark in the interaction instruction. If the interaction instruction is a left fuzzy instruction, the left curtain is drawn according to the interaction instruction, otherwise the right curtain is drawn.

[0104] Obtain the real-time time point, retrieve the corresponding room comprehensive model B in the database according to the time point, identify the coordinates of the virtual irradiation point in the room comprehensive model B, identify the coordinates of the first curtain edge node, connect the virtual irradiation point and the first curtain edge node, identify the first slope of the connection line. If the first slope is greater than the threshold, obtain the right fuzzy instruction, identify the coordinates of the second curtain edge node after adjustment of the right fuzzy instruction, connect the virtual irradiation point and the second curtain edge node, identify the second slope of the connection line, compare with the database, anchor the right fuzzy instruction corresponding to the second slope that meets the threshold, and use the right fuzzy instruction as the first active instruction, otherwise obtain the left fuzzy instruction.

[0105] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprises", "comprising" or any other variation thereof is intended to cover a non-exclusive inclusion, such that a process, method, article or device comprising a series of elements includes not only those elements, but also other elements not expressly listed, or elements inherent to such process, method, article or device.

[0106] Finally, it should be noted that the above are only the preferred embodiments of the present invention and are not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features. 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. An intelligent interaction method based on artificial intelligence, characterized in that: The method comprises: Collecting historical operation data of the user, and analyzing the historical operation data to obtain an adjustment threshold value corresponding to the user's operation habits; Collecting the user's behavior after interacting with the smart home, and analyzing the behavior to obtain the user's behavioral habit characteristics; Analyze the user's voice features, match fuzzy instructions for the voice features shown, obtain a fuzzy instruction set, filter the fuzzy instruction set according to the adjustment threshold, analyze the impact of the user's scene characteristics and the behavioral habit characteristics on the fuzzy instructions and obtain the priority of the fuzzy instructions, and determine the interaction instructions according to the priority of the fuzzy instructions.

2. The intelligent interaction method based on artificial intelligence according to claim 1, characterized in that: The steps of analyzing the user's voice features, matching the fuzzy instructions, further analyzing the scene where the user is located, analyzing the scene environment features, analyzing the impact of the scene environment features on the fuzzy instructions, evaluating the priority of the fuzzy instructions, and determining the interactive instructions include the following steps: Analyze the semantics of the user's voice command, match the corresponding fuzzy command according to the semantics in the user's voice command, and filter the fuzzy command set according to the adjustment threshold; Analyze the three-dimensional data of the room and build a comprehensive model of the room; In the constructed comprehensive room model, the light irradiation area after the curtains are opened under various fuzzy commands is simulated, the influence of the light irradiation area and user behavior habits on the fuzzy commands is analyzed and determined, and the priority of the fuzzy commands is evaluated, and the fuzzy command with the highest priority is selected as the interactive command.

3. The intelligent interaction method based on artificial intelligence according to claim 2, characterized in that: The step of analyzing the semantics of the user's voice command and matching the corresponding fuzzy command according to the semantics in the user's voice command comprises the following steps: Acquire the user's voice command, identify the semantics of the user's voice command, retrieve the command set set in the data and compare them. If the similarity is greater than a threshold, mark the first command with the greatest similarity as an interactive command. Otherwise, extract the keywords in the user's voice command, search the database according to the keywords, retrieve the corresponding fuzzy commands in the database, and obtain a fuzzy command set. Obtain the fuzzy instruction, identify the curtain adjustment distance l3 corresponding to the fuzzy instruction, and compare it with the user's habitual adjustment threshold. If the curtain adjustment distance l3 is greater than the user's habitual adjustment threshold, the fuzzy instruction will be deleted, otherwise the system will continue to detect.

4. The intelligent interaction method based on artificial intelligence according to claim 3, characterized in that: The constructed room comprehensive model simulates the light irradiation area after the curtains are opened under each fuzzy instruction, analyzes and determines the influence of the light irradiation area and user behavior habits on the fuzzy instruction, evaluates the priority of the fuzzy instruction, and determines the interactive instruction, including the following steps: Obtain a fuzzy instruction set, simulate and control the curtains in the room comprehensive model according to the fuzzy instruction set, and output the room model A after the curtains are adjusted according to the fuzzy instructions, identify the edge inflection points c, d, e, and f of the curtains, wherein the edge inflection points c and d represent the upper and lower edge inflection points on the inner side of the left curtain, respectively, and the edge inflection points e and f represent the upper and lower edge inflection points on the inner side of the right curtain, respectively, mark the edge inflection points of the curtains in the room model A, connect the virtual illumination points with the edge inflection points c, d, e, and f, respectively, and construct a connection line between the virtual illumination points and the edge inflection point coordinates; Acquire the visual image in the room, identify the user contour nodes, build the user model, identify the plane where the user model is located, identify the intersection of the line connecting the virtual illumination point and the edge inflection point coordinates with the plane where the user is located, connect adjacent intersections, fit the lines connecting adjacent intersections, build a light radiation model, identify the angle θ1 of adjacent edges of the light radiation model, identify the coordinates of the adjacent intersections, and calculate the lengths l1 and l2 of the lines between the adjacent intersections after fitting using a distance formula, where l1 and l2 represent two adjacent edges in the light radiation model, and calculate the area of ​​the light radiation model using the formula S1=α×sinθ1×l1×l2, where α represents the error coefficient of the area of ​​the light radiation model, and retrieve the corresponding influence coefficient β on the fuzzy instruction in the database according to the area of ​​the light radiation model. i , where, i=1,2,3...n,β i represents the influence coefficient of the area of ​​the light radiation model after the curtain is adjusted by the i-th fuzzy instruction on the fuzzy instruction; Identify the body area of ​​the user who is used to sunbathing, anchor the area of ​​the light radiation model, overlap and compare the areas, mark the user body area in the overlapping area as the first area, identify the area S2 of the first area, and calculate the ratio of the area of ​​the first area to the area of ​​the light radiation model According to the ratio, a database is searched to retrieve the corresponding influence coefficient δ on the fuzzy instruction; Obtain the user's behavioral habit characteristics and identify the user's behavioral habits. If the user is used to sunbathing, the influence weight ε1 of the influence coefficient δ set in the database is retrieved; otherwise, the influence weight ε2 of the influence coefficient δ set in the database is retrieved; Get the remaining fuzzy instructions, get the basic priority score P0 set by the fuzzy instructions, and calculate the priority score P1 of the remaining fuzzy instructions by the formula = (β+δ×ε j )×P0, where j=1,2, the priority scores of the remaining fuzzy instructions are sorted in descending order, the first-ranked fuzzy instruction is selected, and the interactive instruction is generated according to the first-ranked fuzzy instruction.

5. The intelligent interaction method based on artificial intelligence according to claim 4, characterized in that: After determining the interaction instruction, the method further includes the following steps: Analyze user behavior and scene environment changes, match active commands, and interact with smart homes based on active commands and interactive commands; Identify the coordinate changes of the virtual illumination point in the room comprehensive model, identify the changes of the edge nodes during the curtain pulling process, obtain the slope changes of the line connecting the virtual illumination point and the edge node, determine the slope that meets the preset judgment conditions, and then determine the position of the curtain edge node, and determine the active instruction according to the position of the curtain edge node; An interaction instruction is obtained, and a mark in the interaction instruction is identified. If the interaction instruction is a left fuzzy instruction, the left curtain is opened according to the interaction instruction, otherwise the right curtain is opened.

6. An intelligent interactive system based on artificial intelligence, characterized by: The system includes: a data acquisition module, a behavior recognition module, a command analysis module and an interactive control module; The data acquisition module is used to establish an interactive database, collect sensor comprehensive data in real time and enter the data into the interactive database, and enter the user room comprehensive information into the interactive database; The behavior recognition module is used to recognize the historical operation data of the user, analyze the historical operation data to obtain the adjustment threshold corresponding to the user's operation habits, and analyze the behavior to obtain the user's behavior habit characteristics; The instruction analysis module is used to analyze the user's voice features, match fuzzy instructions to the voice features, obtain a fuzzy instruction set, filter the fuzzy instruction set according to the adjustment threshold, analyze the impact of the user's scene characteristics and the behavioral habit characteristics on the fuzzy instructions and obtain the priority of the fuzzy instructions, and determine the interaction instructions according to the priority of the fuzzy instructions.

7. The intelligent interactive system based on artificial intelligence according to claim 6, characterized in that: The data acquisition module includes a sensor module and a room comprehensive information entry module; The sensor module is used to establish an interactive database, collect sensor comprehensive data in real time and enter the data into the interactive database; The room comprehensive information input module is used to input the user room comprehensive information into the interactive database.

8. The intelligent interactive system based on artificial intelligence according to claim 7, characterized in that: The instruction analysis module is also used for: Analyze the semantics of the user's voice command, match the corresponding fuzzy command according to the semantics in the user's voice command, and filter the fuzzy command set according to the adjustment threshold; Analyze the three-dimensional data of the room and build a comprehensive model of the room; In the constructed comprehensive room model, the light irradiation area after the curtains are opened under various fuzzy commands is simulated, the influence of the light irradiation area and user behavior habits on the fuzzy commands is analyzed and determined, and the priority of the fuzzy commands is evaluated, and the fuzzy command with the highest priority is selected as the interactive command.

9. The intelligent interactive system based on artificial intelligence according to claim 8, characterized in that: The instruction analysis module includes a fuzzy instruction matching module, a scenario analysis module and a priority evaluation module; The fuzzy instruction matching module is used to obtain the user's voice instruction, identify the semantics of the user's voice instruction, retrieve the instruction set set in the data and compare them, and if the similarity is greater than a threshold, mark the first instruction with the greatest similarity as an interactive instruction, otherwise extract the keywords in the user's voice instruction, search the database according to the keywords, retrieve the corresponding fuzzy instructions in the database, and obtain the fuzzy instruction set; Obtain the fuzzy instruction, identify the curtain adjustment distance l3 corresponding to the fuzzy instruction, and compare it with the user's habitual adjustment threshold. If the curtain adjustment distance l3 is greater than the user's habitual adjustment threshold, the fuzzy instruction is deleted, otherwise the system continues to detect; The scenario analysis module is used to obtain a fuzzy instruction set, simulate the control of the curtains in the room comprehensive model according to the fuzzy instruction set, and output the room model A after the curtains are adjusted according to the fuzzy instructions, identify the edge inflection points c, d, e, and f of the curtains, wherein the edge inflection points c and d represent the upper and lower edge inflection points on the inner side of the left curtain, respectively, and the edge inflection points e and f represent the upper and lower edge inflection points on the inner side of the right curtain, respectively, mark the edge inflection points of the curtains in the room model A, connect the virtual irradiation points with the edge inflection points c, d, e, and f, respectively, and construct a connection line between the virtual irradiation points and the edge inflection point coordinates; Acquire the visual image in the room, identify the user contour nodes, build the user model, identify the plane where the user model is located, identify the intersection of the line connecting the virtual illumination point and the edge inflection point coordinates with the plane where the user is located, connect adjacent intersections, fit the lines connecting adjacent intersections, build a light radiation model, identify the angle θ1 of adjacent edges of the light radiation model, identify the coordinates of the adjacent intersections, and calculate the lengths l1 and l2 of the lines between the adjacent intersections after fitting using a distance formula, where l1 and l2 represent two adjacent edges in the light radiation model, and calculate the area of ​​the light radiation model using the formula S1=α×sinθ1×l1×l2, where α represents the error coefficient of the area of ​​the light radiation model, and retrieve the corresponding influence coefficient β on the fuzzy instruction in the database according to the area of ​​the light radiation model. i , where, i=1,2,3...n,β i represents the influence coefficient of the area of ​​the light radiation model after the curtain is adjusted by the i-th fuzzy instruction on the fuzzy instruction; Identify the body area of ​​the user who is used to sunbathing, anchor the area of ​​the light radiation model, overlap and compare the areas, mark the user body area in the overlapping area as the first area, identify the area S2 of the first area, and calculate the ratio of the area of ​​the first area to the area of ​​the light radiation model According to the ratio, a database is searched to retrieve the corresponding influence coefficient δ on the fuzzy instruction; Obtain the user's behavioral habit characteristics and identify the user's behavioral habits. If the user is used to sunbathing, the influence weight ε1 of the influence coefficient δ set in the database is retrieved; otherwise, the influence weight ε2 of the influence coefficient δ set in the database is retrieved; The priority evaluation module is used to obtain the remaining fuzzy instructions, obtain the basic priority score P0 set by the fuzzy instructions, calculate the priority score of the remaining fuzzy instructions by the formula P1=β×δ×P0, sort the priority scores of the remaining fuzzy instructions in descending order, select the first-ranked fuzzy instructions, and generate interactive instructions according to the first-ranked fuzzy instructions.

10. The intelligent interactive system based on artificial intelligence according to claim 9, characterized in that: The interactive control module is also used to identify the coordinate changes of the virtual illumination point in the room comprehensive model, identify the changes of the edge nodes during the curtain pulling process, obtain the slope changes of the line connecting the virtual illumination point and the edge node, determine the slope that meets the preset judgment conditions, and then determine the position of the curtain edge node and determine the active instruction; An interaction instruction is obtained, and a mark in the interaction instruction is identified. If the interaction instruction is a left fuzzy instruction, the left curtain is opened according to the interaction instruction, otherwise the right curtain is opened.

Images