Intelligent furniture comfort adjustment method and system based on behavioral feature recognition

Abstract

The present invention discloses a method and system for adjusting the comfort of intelligent furniture based on behavioral feature recognition. By constructing a first correlation matrix and a second correlation matrix, the equipment parameters, environmental parameters and physiological parameters that have the greatest impact on office efficiency and comfort are respectively screened out, thereby realizing the organic integration of multi-dimensional parameters. Furthermore, the user behavior characteristics are incorporated into the construction of an office efficiency model and a comfort model as key variables, so that the model can more accurately reflect the actual needs of users and improve the level of personalized services. Finally, an office behavior adaptation model is constructed in combination with the office efficiency model and the comfort model, thereby realizing the dual optimization of office efficiency and user comfort and providing a more intelligent and comprehensive equipment adjustment strategy. The intelligent furniture can automatically adjust to the optimal state according to the real-time behavioral characteristics of the user, thereby not only improving office efficiency, but also ensuring that the user is in the most comfortable working environment, thereby realizing an intelligent and personalized office experience.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method and system for adjusting the comfort level of intelligent furniture based on behavioral feature recognition. Background Art

[0002] Smart furniture is increasingly used in office environments to improve office efficiency and employee comfort; existing technologies mainly focus on adjusting a single parameter, such as adjusting the air conditioning temperature according to the ambient temperature, or adjusting the seat height according to the user's preferences. However, existing methods ignore the complex interactive relationship between user behavior characteristics and environmental, physiological and equipment parameters, resulting in limited adjustment effects of smart furniture, which cannot fully meet personalized needs, and cannot find the best balance between improving office efficiency and ensuring user comfort.

[0003] The above information disclosed in the Background section is only for enhancement of understanding of the background of the present application and therefore it may contain information that does not constitute the prior art that is already known to a person of ordinary skill in the art. Summary of the invention

[0004] In response to the above-mentioned deficiencies, the present application proposes a method and system for adjusting the comfort of intelligent furniture based on behavioral feature recognition. By constructing a first correlation matrix and a second correlation matrix, the equipment parameters, environmental parameters and physiological parameters that have the greatest impact on office efficiency and comfort are respectively screened out to achieve an organic integration of multi-dimensional parameters. Furthermore, user behavior characteristics are incorporated into the construction of office efficiency models and comfort models as key variables, so that the models can more accurately reflect the actual needs of users and improve the level of personalized services. Finally, an office behavior adaptation model is constructed by combining the office efficiency model and the comfort model to achieve dual optimization of office efficiency and user comfort, and provide a more intelligent and comprehensive equipment adjustment strategy. Intelligent furniture can automatically adjust to the optimal state according to the user's real-time behavioral characteristics, which not only improves office efficiency, but also ensures that users are in the most comfortable working environment, thereby achieving an intelligent and personalized office experience.

[0005] In a first aspect, a technical solution provided in an embodiment of the present invention is a method for adjusting the comfort level of smart furniture based on behavioral feature recognition, comprising the following steps:

[0006] Obtain a first correlation matrix of device parameter information and environmental parameter information relative to an office efficiency factor, and obtain a first valid device parameter information set and an environmental parameter information set according to the first correlation matrix; and construct an office efficiency model based on the first valid device parameter information set and the environmental parameter information set in combination with user behavior characteristics;

[0007] Obtain a second correlation matrix of device parameter information and physiological parameter information relative to the comfort factor, and obtain a second valid device parameter information set and a physiological parameter information set according to the second correlation matrix; and construct a comfort model based on the second valid device parameter information set and the physiological parameter information set in combination with user behavior characteristics;

[0008] Build an office behavior adaptation model based on the office efficiency model and comfort model; identify real-time behavior characteristics and output device adjustment strategies based on the office behavior adaptation model, and control smart furniture to perform corresponding actions based on the device adjustment strategies.

[0009] In this solution, by constructing a correlation matrix, this application can effectively screen out equipment parameters, environmental parameters and physiological parameters that have a significant impact on office efficiency and comfort, reduce the data processing burden, and improve system efficiency; incorporate user behavior characteristics so that the constructed office efficiency model and comfort model are closer to the actual needs of users, thereby generating more personalized equipment adjustment strategies and improving user experience; by constructing an office behavior adaptation model, the two dimensions of office efficiency and comfort are comprehensively considered to achieve comprehensive optimization of the smart furniture system, which not only improves work efficiency but also ensures user comfort; real-time identification of user behavior characteristics, combined with the office behavior adaptation model, and dynamic output of equipment adjustment strategies enable smart furniture to respond to user needs immediately and create a more suitable office environment.

[0010] As a preferred solution of the present application, the step of obtaining a first correlation matrix of device parameter information and environmental parameter information relative to an office efficiency factor, and obtaining a first valid device parameter information set and an environmental parameter information set according to the first correlation matrix comprises the following steps:

[0011] The office efficiency value is calculated based on the efficiency ratio, quality coefficient, and completion degree of the project on a time scale, and the office efficiency value is binned to obtain several office efficiency intervals, and each office efficiency interval is used as an office efficiency factor;

[0012] Extract the equipment parameter information and environmental parameter information corresponding to each office efficiency factor on a time scale;

[0013] On a time scale, the correlation between the equipment parameter information and the office efficiency factor is calculated according to the correlation algorithm to obtain a first equipment parameter correlation matrix; the elements with correlation values ​​less than a set threshold a1 in the first equipment parameter correlation matrix are removed to obtain a first effective equipment correlation matrix;

[0014] Synchronously, the correlation between the environmental parameter information and the office efficiency factor is calculated on a time scale according to a correlation algorithm to obtain an environmental parameter correlation matrix; the elements in the environmental parameter correlation matrix whose correlation values ​​are less than a set threshold a2 are removed to obtain an effective environmental parameter correlation matrix; wherein the first correlation matrix includes a first effective device correlation matrix and an effective environmental parameter correlation matrix;

[0015] The first effective device parameter information set and the environmental parameter information set belonging to the office efficiency factor are obtained by respectively associating the rows or columns of the first effective device correlation matrix and the effective environmental parameter correlation matrix through the office efficiency factor.

[0016] In this scheme, by comprehensively considering multi-dimensional factors such as efficiency ratio, quality coefficient, completion degree, etc., and binning the office efficiency values, a more accurate office efficiency evaluation result is obtained; by calculating the correlation between equipment parameters and environmental parameters and office efficiency factors, and eliminating elements with correlation values ​​less than the set threshold, the parameters that have a significant impact on office efficiency are effectively screened out; the time scale is introduced into the correlation calculation, and the change of the correlation between parameters and office efficiency over time is dynamically considered to improve the accuracy and refinement of the correlation calculation; by associating the office efficiency factor with the effective equipment parameter correlation matrix and the effective environmental parameter correlation matrix, the effective equipment parameter information set and environmental parameter information set belonging to different office efficiency factors are successfully extracted; the present application can dynamically adjust the parameter set and the office efficiency evaluation model according to the actual scene data, so as to improve the adaptability and intelligence level of the smart furniture system.

[0017] As a preferred solution of the present application, the office efficiency model is constructed based on the first valid device parameter information set and the environment parameter information set in combination with user behavior characteristics; comprising the following steps:

[0018] Using the first valid device parameter information set and the environmental parameter information set as input data of the first machine learning model, using the corresponding office efficiency factor as output data of the first machine learning model, and sequentially training the first machine learning model to obtain a first weight coefficient value; constructing a first input parameter based on the first weight coefficient value, the first valid device parameter information set, and the environmental parameter information set;

[0019] Obtaining user behavior characteristics and a first weight coefficient value on the same time scale as input data of a second machine learning model; using a corresponding office efficiency factor as output data of the second machine learning model, and sequentially training the second machine learning model to obtain a second weight coefficient value; constructing a second input parameter according to the second weight coefficient value, the user behavior characteristics, and the first input parameter;

[0020] The elements in the second input parameter are used as the input of the third machine learning model, and the office efficiency factor is used as the output data of the third machine learning model to construct an office efficiency model, and the third weight coefficient value is obtained by training the office efficiency model.

[0021] In this solution, in order to address the problems of improper selection of model input parameters and weak model generalization ability, the first effective device parameter information set and environmental parameter information set are screened as input data for the first machine learning model to ensure that the input parameters are directly related to office efficiency and improve the accuracy of model prediction; user behavior characteristics are used as input data for the second machine learning model so that the model can fully consider the impact of user behavior on office efficiency and improve the prediction effect and personalization of the model; the first, second and third machine learning models are trained in sequence to obtain accurate weight coefficient values ​​to ensure that the model is fully trained and improve the prediction accuracy and stability of the model; because the model considers multiple dimensions such as device parameters, environmental parameters and user behavior characteristics and has been fully trained, it has strong generalization ability and can adapt to changes in different scenarios or data; according to changes in actual scenario data, the model's input parameters and weight coefficient values ​​can be dynamically adjusted so that the model can continuously adapt to new data and environments and maintain the accuracy of the prediction effect.

[0022] As a preferred solution of the present application, obtaining a second correlation matrix of device parameter information and physiological parameter information relative to the comfort factor, and obtaining a second valid device parameter information set and a physiological parameter information set according to the second correlation matrix; comprising the following steps:

[0023] The comfort value is calculated according to the changes in physiological indicators and behavioral characteristics during the execution of the project on a time scale, and the comfort value is binned to obtain several comfort intervals, and each comfort interval is used as a comfort factor;

[0024] Extract the equipment parameter information and environmental parameter information corresponding to each comfort factor on a time scale;

[0025] On a time scale, the correlation between the equipment parameter information and the comfort factor is calculated according to the correlation algorithm to obtain a second equipment parameter correlation matrix; the elements with correlation values ​​less than a set threshold a3 in the second equipment parameter correlation matrix are removed to obtain a second effective equipment correlation matrix;

[0026] Synchronously, the correlation between the physiological parameter information and the comfort factor is calculated on a time scale according to a correlation algorithm to obtain a physiological parameter correlation matrix; the elements in the physiological parameter correlation matrix whose correlation values ​​are less than a set threshold a4 are removed to obtain an effective physiological parameter correlation matrix; wherein the second correlation matrix includes a second effective device correlation matrix and an effective physiological parameter correlation matrix;

[0027] The rows or columns of the second effective device correlation matrix and the effective physiological parameter correlation matrix are respectively associated with the comfort factor to obtain the second effective device parameter information set and the physiological parameter information set belonging to the comfort factor.

[0028] In this scheme, since traditional comfort evaluation often only focuses on a single physiological index or environmental parameter, such as temperature and humidity, and ignores the impact of the user's physiological and behavioral changes on comfort on a time scale, this scheme calculates the comfort value by comprehensively considering the changes in the user's physiological indexes and behavioral characteristics during the project execution, and obtains the comfort factor by binning. This makes the comfort evaluation more comprehensive and multi-dimensional, and can better reflect the user's true feelings; among a large number of equipment parameters and physiological parameters, many parameters have no direct correlation with comfort or have a very low correlation. These redundant parameters not only increase the complexity of data processing, but may also affect the accuracy of the comfort model; this scheme effectively screens out the parameter set that has a significant impact on comfort by calculating the correlation between equipment parameters and physiological parameters and comfort factors, and eliminating elements with correlation values ​​less than the set threshold. This reduces the complexity of data processing and improves the accuracy of the comfort model. Traditional comfort models often only consider static parameter relationships, while ignoring the dynamic changes in the correlation between parameters and comfort; this solution introduces a time scale in the correlation calculation, and dynamically considers the changes in the correlation between parameters and comfort over time. This makes the comfort model more consistent with the dynamic changes in actual scenarios, and improves the accuracy and adaptability of the model. Due to the large number of parameters and the complexity of correlation calculation, it is a technical challenge to effectively extract parameter sets that have a significant impact on comfort; this solution successfully extracts effective equipment parameter information sets and physiological parameter information sets belonging to different comfort factors by associating comfort factors with effective equipment parameter correlation matrices and effective physiological parameter correlation matrices, providing strong support for building a more accurate comfort model; by overcoming technical problems such as the lack of multi-dimensional considerations in comfort evaluation, low correlation between parameter redundancy and comfort, lack of dynamics in comfort models, and difficulty in extracting effective parameter sets, a more comprehensive, accurate and dynamic comfort evaluation is achieved in actual scenarios.

[0029] As a preferred embodiment of the present application, the physiological indicators include skin temperature, heart rate, blood pressure and sweating amount; the behavioral characteristic changes include posture adjustment degree, activity level change, and facial expression tension;

[0030] The comfort value is obtained by weighted summation based on skin temperature, heart rate, blood pressure, sweating amount, posture adjustment degree, activity level change, facial expression tension and corresponding weight coefficients.

[0031] In this scheme, since traditional comfort evaluation often only focuses on a single physiological index, such as skin temperature or heart rate, while ignoring other equally important physiological and behavioral characteristics; this scheme comprehensively considers multiple physiological indicators such as skin temperature, heart rate, blood pressure, and sweating, and combines the changes in behavioral characteristics such as posture adjustment degree, activity level, and facial expression tension, and performs weighted summation to obtain the comfort value. This makes the comfort evaluation more comprehensive and can more accurately reflect the user's true feelings. When determining the contribution of each physiological index and behavioral characteristic change to the comfort value, this scheme determines the weight coefficient corresponding to each physiological index and behavioral characteristic change through the hierarchical analysis method, thereby ensuring the objectivity and accuracy of the weight coefficient; traditional comfort evaluation often only considers static physiological indicators and behavioral characteristics, while ignoring the dynamic changes of these indicators and characteristics on the time scale. This scheme obtains the real-time comfort value by collecting the changes in the user's physiological indicators and behavioral characteristics in real time, and performing dynamic weighted summation. This makes the comfort evaluation more in line with the dynamic changes in the actual scene, and can more accurately and comprehensively reflect the user's comfort feeling at different time points.

[0032] As a preferred solution of the present application, a change comfort model is constructed based on the second effective device parameter information set and the physiological parameter information set combined with user behavior characteristics; the steps include:

[0033] Using the second valid device parameter information set and the physiological parameter information set as input data of the fourth machine learning model, using the corresponding comfort factor as output data of the fourth machine learning model, and sequentially training the fourth machine learning model to obtain a fourth weight coefficient value; constructing a third input parameter according to the fourth weight coefficient value, the second valid device parameter information set, and the physiological parameter information set;

[0034] Obtaining user behavior characteristics and a third weight coefficient value on the same time scale as input data of a fifth machine learning model; using the corresponding comfort factor as output data of the fifth machine learning model, and sequentially training the fifth machine learning model to obtain a fifth weight coefficient value; constructing a fourth input parameter according to the fifth weight coefficient value, the user behavior characteristics, and the third input parameter;

[0035] Use the elements in the fourth input parameter as the input of the sixth machine learning model, and use the comfort factor as the output data of the sixth machine learning model to construct an office efficiency model, and obtain the sixth weight coefficient value by training the office efficiency model.

[0036] In this solution, the traditional comfort model often only focuses on a single physiological parameter or device parameter, while ignoring the influence of user behavior characteristics on comfort, resulting in inaccurate model prediction results; this solution constructs a variable comfort model by comprehensively considering the second effective device parameter information set, the physiological parameter information set and the user behavior characteristics. In this way, the model can more comprehensively reflect the user's perception of comfort and improve the accuracy and personalization of the prediction. When constructing a comfort model, if the selected training data has no direct correlation with comfort or has a low correlation, the model training effect will be poor, which will in turn affect the accuracy of the prediction; this solution selects the second effective device parameter information set and the physiological parameter information set as the input data of the fourth machine learning model, and trains the corresponding comfort factor as the output data. It can ensure that the training data is directly related to comfort and improve the prediction effect of the model; if the comfort model is only trained for specific scenarios or data, the generalization ability will be weak and it will be difficult to adapt to changes in different scenarios or data; this solution constructs a variable comfort model with strong generalization ability by training the fourth, fifth and sixth machine learning models in sequence and obtaining the corresponding weight coefficient values. The model can adapt to different scenarios or data changes and maintain the accuracy of the prediction effect. If the data used in the model training process is not comprehensive or the number of training times is insufficient, the model weight coefficient will be inaccurate, which will affect the prediction effect of the model. This solution fully trains the fourth, fifth and sixth machine learning models to ensure accurate weight coefficient values, so that the variable comfort model is fully trained and the prediction accuracy and stability of the model are improved.

[0037] As a preferred solution of the present application, an office behavior adaptation model is constructed according to an office efficiency model and a comfort model; the real-time behavior characteristics are identified and combined with the office behavior adaptation model to output a device adjustment strategy, and the smart furniture is controlled to perform corresponding actions based on the device adjustment strategy; the steps include:

[0038] The office efficiency factor, the comfort factor, the third weight coefficient value, the sixth weight coefficient value, the physiological parameter information, and the user behavior characteristics are used as inputs of the seventh machine learning model; the equipment parameter information and the environmental parameter information are used as outputs of the seventh machine learning model to construct an office behavior adaptation model;

[0039] Acquire real-time user behavior feature data, and lock the office efficiency factor with the highest office efficiency and its corresponding third weight coefficient value, the comfort factor with the highest comfort value and its corresponding sixth weight coefficient value, and obtain target device parameter information and target environment parameter information in combination with the office behavior adaptation model;

[0040] A device adjustment strategy is formulated according to target device parameter information and target environment parameter information, and the smart furniture is controlled to perform corresponding actions based on the device adjustment strategy.

[0041] In this solution, in the traditional office environment setting, furniture and equipment are often fixed and cannot be intelligently adjusted according to the user's real-time behavior characteristics and comfort needs, resulting in poor adaptability between the office environment and user behavior, affecting office efficiency and user comfort. This solution can identify the user's real-time behavior characteristics by building an office behavior adaptation model, and formulate equipment adjustment strategies based on the output of the office efficiency model and the comfort model, thereby controlling the smart furniture to perform corresponding actions; the office environment can be intelligently adjusted according to the user's real-time needs and comfort feelings, improving the adaptability of the office environment to user behavior. When formulating equipment adjustment strategies, there is often a lack of scientific basis, and adjustments are made only based on experience or preset rules, resulting in poor adjustment effects and failure to meet the actual needs of users. This solution uses the seventh machine learning model to construct an office behavior adaptation model, taking office efficiency factors, comfort factors, weight coefficient values, physiological parameter information, user behavior characteristics, etc. as inputs, and equipment parameter information and environmental parameter information as outputs. When formulating equipment adjustment strategies, the comprehensive impact of various factors can be fully considered to ensure the scientificity and effectiveness of the adjustment strategies. Smart furniture control lacks intelligence and personalization. Traditional smart furniture control is often just a simple switch operation or switching of preset modes, lacking intelligent and personalized control strategies, and unable to perform intelligent adjustments based on users' real-time needs and comfort levels. This solution identifies users' real-time behavior characteristics and combines the output of the office behavior adaptation model to formulate personalized equipment adjustment strategies, thereby controlling smart furniture to perform corresponding actions, making the control of smart furniture more intelligent and personalized, and able to meet users' real-time needs and comfort levels.

[0042] In a second aspect, a technical solution provided in an embodiment of the present invention is: a smart furniture comfort adjustment system, comprising:

[0043] A first calculation module: obtaining a first correlation matrix of device parameter information and environmental parameter information relative to an office efficiency factor, and obtaining a first valid device parameter information set and an environmental parameter information set according to the first correlation matrix;

[0044] The first construction module: constructs an office efficiency model based on the first valid device parameter information set and the environment parameter information set combined with user behavior characteristics;

[0045] The second calculation module is used to obtain a second correlation matrix of the device parameter information and the physiological parameter information relative to the comfort factor, and obtain a second valid device parameter information set and a physiological parameter information set according to the second correlation matrix;

[0046] The second building module: building a comfort model based on the second effective device parameter information set and the physiological parameter information set combined with user behavior characteristics;

[0047] The second building module: building an office behavior adaptation model based on the office efficiency model and the comfort model;

[0048] Execution model: Identify real-time behavior characteristics and combine them with the office behavior adaptation model to output device adjustment strategies, and control smart furniture to perform corresponding actions based on the device adjustment strategies.

[0049] In a third aspect, a technical solution provided in an embodiment of the present invention is an electronic device, comprising a memory and a processor, wherein a computer program is stored in the memory, and when the processor calls the computer program in the memory, the steps of the smart furniture comfort adjustment method based on behavioral feature recognition are implemented.

[0050] In a fourth aspect, a technical solution provided in an embodiment of the present invention is a storage medium, in which computer executable instructions are stored. When the computer executable instructions are loaded and executed by a processor, the steps of the smart furniture comfort adjustment method based on behavioral feature recognition are implemented.

[0051] Beneficial effects of the present invention: The present application proposes a method for adjusting the comfort of intelligent furniture based on behavioral feature recognition. By constructing a first correlation matrix and a second correlation matrix, the equipment parameters, environmental parameters and physiological parameters that have the greatest impact on office efficiency and comfort are respectively screened out, thereby realizing the organic integration of multi-dimensional parameters. Furthermore, user behavior characteristics are incorporated into the construction of office efficiency models and comfort models as key variables, so that the models can more accurately reflect the actual needs of users and improve the level of personalized services. Finally, an office behavior adaptation model is constructed by combining the office efficiency model and the comfort model to achieve dual optimization of office efficiency and user comfort, and provide a more intelligent and comprehensive equipment adjustment strategy. Intelligent furniture can automatically adjust to the optimal state according to the real-time behavioral characteristics of users, which not only improves office efficiency, but also ensures that users are in the most comfortable working environment, thereby realizing an intelligent and personalized office experience.

[0052] The above invention content is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented according to the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Other features, objects and advantages of the present invention will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings. The drawings are only for the purpose of illustrating preferred embodiments and are not to be considered as limiting the present invention. Also, the same reference symbols are used throughout the drawings to represent the same parts.

[0054] Figure 1 The present invention is a flow chart of the method for adjusting the comfort level of intelligent furniture based on behavioral feature recognition. DETAILED DESCRIPTION

[0055] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific implementation method described herein is only an optimal embodiment of the present invention, which is only used to explain the present invention and does not limit the scope of protection of the present invention. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0056] Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flow charts. Although the flow charts describe the operations (or steps) as sequential processes, many of the operations (or steps) therein can be implemented in parallel, concurrently, or simultaneously. In addition, the order of the operations can be rearranged. The process can be terminated when its operation is completed, but can also have additional steps not included in the drawings; the process can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0057] Example: Figure 1 As shown, the intelligent furniture comfort adjustment method based on behavioral feature recognition includes the following steps:

[0058] S1. Obtain a first correlation matrix of device parameter information and environmental parameter information relative to an office efficiency factor, and obtain a first valid device parameter information set and an environmental parameter information set according to the first correlation matrix; and construct an office efficiency model based on the first valid device parameter information set and the environmental parameter information set in combination with user behavior characteristics.

[0059] Specifically, the method of obtaining a first correlation matrix of device parameter information and environmental parameter information relative to an office efficiency factor, and obtaining a first valid device parameter information set and an environmental parameter information set according to the first correlation matrix, comprises the following steps:

[0060] The office efficiency value is calculated based on the efficiency ratio, quality coefficient, and completion degree of the project on a time scale, and the office efficiency value is binned to obtain several office efficiency intervals, and each office efficiency interval is used as an office efficiency factor;

[0061] Extract the equipment parameter information and environmental parameter information corresponding to each office efficiency factor on a time scale;

[0062] On a time scale, the correlation between the equipment parameter information and the office efficiency factor is calculated according to the correlation algorithm to obtain a first equipment parameter correlation matrix; the elements with correlation values ​​less than a set threshold a1 in the first equipment parameter correlation matrix are removed to obtain a first effective equipment correlation matrix;

[0063] Synchronously, the correlation between the environmental parameter information and the office efficiency factor is calculated on a time scale according to a correlation algorithm to obtain an environmental parameter correlation matrix; the elements in the environmental parameter correlation matrix whose correlation values ​​are less than a set threshold a2 are removed to obtain an effective environmental parameter correlation matrix; wherein the first correlation matrix includes a first effective device correlation matrix and an effective environmental parameter correlation matrix;

[0064] The first effective device parameter information set and the environmental parameter information set belonging to the office efficiency factor are obtained by respectively associating the rows or columns of the first effective device correlation matrix and the effective environmental parameter correlation matrix through the office efficiency factor.

[0065] In this embodiment, the office efficiency value is calculated based on the efficiency ratio, quality coefficient, and completion of the project on a time scale. For example, the efficiency ratio is measured by the ratio of the task completion time to the expected time, the quality coefficient is evaluated by the accuracy and satisfaction of the task completion, and the completion degree is the completion percentage of the task. The efficiency ratio, quality coefficient, and completion degree are weighted by the hierarchical analysis method; then, an office efficiency value is obtained by weighted summation based on the efficiency ratio, quality coefficient, completion degree, and their corresponding weight values; the office efficiency value is binned and divided into several intervals, each of which represents an office efficiency factor. For example, the office efficiency value can be divided into intervals such as "low efficiency", "medium efficiency", and "high efficiency"; for each office efficiency factor, the corresponding device parameter information and environmental parameter information on a time scale are extracted. The device parameter information includes various settings and states of smart furniture, such as seat height, table height, tilt angle, light brightness, screen angle, etc.; the environmental parameter information includes the temperature, humidity, light intensity, noise level, etc. in the office.

[0066] Furthermore, on a time scale, the correlation between the device parameter information and the office efficiency factor is calculated according to a correlation algorithm (such as the Pearson correlation coefficient, the Spearman rank correlation coefficient, etc.) to obtain a first device parameter correlation matrix. Similarly, the correlation between the environmental parameter information and the office efficiency factor is calculated to obtain an environmental parameter correlation matrix. The elements whose correlation values ​​in the first device parameter correlation matrix are less than the set threshold a1 are removed to obtain a first effective device correlation matrix. Similarly, the elements whose correlation values ​​in the environmental parameter correlation matrix are less than the set threshold a2 are removed to obtain an effective environmental parameter correlation matrix. Among them, a1 and a2 are manually set according to needs; for example, a1=a2=0.3.

[0067] Furthermore, the first effective device parameter information set and the environmental parameter information set belonging to the office efficiency factor are respectively associated with the rows or columns of the first effective device correlation matrix and the effective environmental parameter correlation matrix through the office efficiency factor. For example, for the office efficiency factor "high efficiency", the corresponding effective device parameter information set may include "seat height", "tilt angle", "screen tilt angle", etc.; the corresponding effective environmental parameter information set may include "temperature", "humidity", "light brightness", etc.; The input of the model includes the first effective device parameter information set, the environmental parameter information set and the user behavior characteristics. The user behavior characteristic data is collected through cameras, sensors, etc., including the user's posture, movement, facial expression, activity mode and other behavioral characteristics. Select a suitable machine learning algorithm to build an office efficiency model. The machine learning model can use a neural network algorithm as the model kernel. Use historical data to train the model; use the first effective device parameter information set, the environmental parameter information set and the user behavior characteristics as inputs, and use the corresponding office efficiency value as output to train the model. Through training, the model can learn the relationship between device parameters, environmental parameters and user behavior characteristics and office efficiency. Evaluate the trained model to check its accuracy and generalization ability. Apply the trained office efficiency model to actual scenarios. By collecting device parameter information, environmental parameter information, and user behavior characteristics in real time and inputting them into the model, the current office efficiency value can be predicted, and corresponding adjustments and optimizations can be made based on the prediction results to improve office efficiency. In summary, by obtaining the first correlation matrix and the effective parameter information set, and combining user behavior characteristics to build an office efficiency model, it is possible to achieve intelligent management and optimization of the office environment and improve office efficiency.

[0068] Specifically, the method of constructing an office efficiency model based on the first valid device parameter information set and the environment parameter information set in combination with user behavior characteristics comprises the following steps:

[0069] Using the first valid device parameter information set and the environmental parameter information set as input data of the first machine learning model, using the corresponding office efficiency factor as output data of the first machine learning model, and sequentially training the first machine learning model to obtain a first weight coefficient value; constructing a first input parameter based on the first weight coefficient value, the first valid device parameter information set, and the environmental parameter information set;

[0070] Obtaining user behavior characteristics and a first weight coefficient value on the same time scale as input data of a second machine learning model; using a corresponding office efficiency factor as output data of the second machine learning model, and sequentially training the second machine learning model to obtain a second weight coefficient value; constructing a second input parameter according to the second weight coefficient value, the user behavior characteristics, and the first input parameter;

[0071] The elements in the second input parameter are used as the input of the third machine learning model, and the office efficiency factor is used as the output data of the third machine learning model to construct an office efficiency model, and the third weight coefficient value is obtained by training the office efficiency model.

[0072] In this embodiment, firstly, a first machine learning model is constructed and a first weight coefficient value is obtained, and input data is prepared: a first valid device parameter information set: {seat height, tilt angle, light brightness} (assuming that the screen angle is not highly correlated with office efficiency and has been eliminated in the previous step); an environmental parameter information set: {temperature, humidity, light intensity} (assuming that the noise level is not highly correlated with office efficiency and has been eliminated in the previous step); output data is prepared: corresponding office efficiency factors: {low efficiency, medium efficiency, high efficiency}; training the first machine learning model: using historical data, taking the first valid device parameter information set and the environmental parameter information set as input, and the corresponding office efficiency factor as output, the first machine learning model (such as a logistic regression model or a neural network model) is trained; through training, the model learns the relationship between device parameters and environmental parameters and office efficiency, and outputs the first weight coefficient value; constructing the first input parameter: constructing the first input parameter according to the first weight coefficient value, the first valid device parameter information set, and the environmental parameter information set. For example, the first input parameter may be a vector containing weighted seat height, tilt angle, light brightness, temperature, humidity, and light intensity.

[0073] Secondly, construct a second machine learning model and obtain a second weight coefficient value; prepare input data: user behavior characteristics: {posture, movement frequency, facial expression, activity mode}; first weight coefficient value: obtained from the first machine learning model; prepare output data: corresponding office efficiency factor: {low efficiency, medium efficiency, high efficiency}; train the second machine learning model: use historical data, user behavior characteristics and the first weight coefficient value as input, and the corresponding office efficiency factor as output to train the second machine learning model (such as a support vector machine model). Through training, the model learns the comprehensive relationship between user behavior characteristics, device parameters, environmental parameters and office efficiency, and outputs the second weight coefficient value. Construct the second input parameter: construct the second input parameter based on the second weight coefficient value, user behavior characteristics, and the first input parameter. For example, the second input parameter may be a vector containing weighted posture, movement frequency, facial expression, activity mode, and each element in the first input parameter.

[0074] Finally, build an office efficiency model and obtain the third weight coefficient value; prepare input data: elements in the second input parameter: including weighted user behavior characteristics, device parameters and environmental parameters. Prepare output data: corresponding office efficiency factors: {low efficiency, medium efficiency, high efficiency}; train the third machine learning model: use historical data, take the second input parameter as input, and the corresponding office efficiency factor as output to train the third machine learning model (such as a neural network model). Through training, the model learns the comprehensive relationship between user behavior characteristics, device parameters, environmental parameters and office efficiency, and outputs the third weight coefficient value. Build an office efficiency model: build an office efficiency model based on the third weight coefficient value and the second input parameter. This model can receive user behavior characteristics, device parameters and environmental parameters as input in real time, and output the current office efficiency factor; in summary, by building three machine learning models and obtaining corresponding weight coefficient values, an office efficiency model that comprehensively considers user behavior characteristics, device parameters and environmental parameters can be built. This model can help the system understand the office efficiency status in real time and adjust and optimize according to actual conditions.

[0075] S2. Obtain a second correlation matrix of device parameter information and physiological parameter information relative to the comfort factor, and obtain a second valid device parameter information set and a physiological parameter information set according to the second correlation matrix; and construct a comfort model based on the second valid device parameter information set and the physiological parameter information set in combination with user behavior characteristics.

[0076] Specifically, obtaining a second correlation matrix of device parameter information and physiological parameter information relative to the comfort factor, and obtaining a second valid device parameter information set and a physiological parameter information set according to the second correlation matrix; comprising the following steps:

[0077] The comfort value is calculated according to the changes in physiological indicators and behavioral characteristics during the execution of the project on a time scale, and the comfort value is binned to obtain several comfort intervals, and each comfort interval is used as a comfort factor;

[0078] Extract the equipment parameter information and environmental parameter information corresponding to each comfort factor on a time scale;

[0079] On a time scale, the correlation between the equipment parameter information and the comfort factor is calculated according to the correlation algorithm to obtain a second equipment parameter correlation matrix; the elements with correlation values ​​less than a set threshold a3 in the second equipment parameter correlation matrix are removed to obtain a second effective equipment correlation matrix;

[0080] Synchronously, the correlation between the physiological parameter information and the comfort factor is calculated on a time scale according to a correlation algorithm to obtain a physiological parameter correlation matrix; the elements in the physiological parameter correlation matrix whose correlation values ​​are less than a set threshold a4 are removed to obtain an effective physiological parameter correlation matrix; wherein the second correlation matrix includes a second effective device correlation matrix and an effective physiological parameter correlation matrix;

[0081] The rows or columns of the second effective device correlation matrix and the effective physiological parameter correlation matrix are respectively associated with the comfort factor to obtain the second effective device parameter information set and the physiological parameter information set belonging to the comfort factor.

[0082] In this embodiment, first, the second correlation matrix of the device parameter information and the physiological parameter information relative to the comfort factor is obtained; the specific process is as follows: Calculate the comfort value: Calculate the comfort value according to the changes in the physiological indicators (such as heart rate, blood pressure) and behavioral characteristics (such as posture, movement frequency) during the execution of the project on a time scale. For example, the weighted average method can be used to assign different weights to the changes in the physiological indicators and behavioral characteristics, and then sum them to obtain the comfort value. Binning to obtain the comfort interval: Binning the comfort value to divide it into several comfort intervals. For example, the comfort value can be divided into three intervals of "low comfort", "medium comfort" and "high comfort". Extract the corresponding device parameter information and physiological parameter information: For each comfort interval (i.e., comfort factor), extract the corresponding device parameter information (such as seat height, tilt angle) and physiological parameter information (such as heart rate, blood pressure) on a time scale.

[0083] Secondly, calculate the correlation matrix and obtain the valid parameter information set; the specific process is: Calculate the device parameter correlation matrix: On the time scale, calculate the correlation between the device parameter information and the comfort factor respectively according to the correlation algorithm (such as the Pearson correlation coefficient) to obtain the second device parameter correlation matrix. For example, the correlation between the seat height and "low comfort", "medium comfort" and "high comfort" can be calculated. Eliminate elements with correlation values ​​less than the set threshold: Eliminate elements with correlation values ​​less than the set threshold a3 in the second device parameter correlation matrix to obtain the second valid device correlation matrix. For example, if the set threshold a3 is 0.5, then the elements with correlation values ​​less than 0.5 will be eliminated. Calculate the physiological parameter correlation matrix: Synchronously, calculate the correlation between the physiological parameter information and the comfort factor respectively according to the correlation algorithm on the time scale to obtain the physiological parameter correlation matrix. For example, the correlation between the heart rate and "low comfort", "medium comfort" and "high comfort" can be calculated. Eliminate elements with correlation values ​​less than the set threshold: Eliminate elements with correlation values ​​less than the set threshold a4 in the physiological parameter correlation matrix to obtain a valid physiological parameter correlation matrix. For example, if the set threshold a4 is 0.6, then elements with correlation values ​​less than 0.6 will be eliminated.

[0084] Finally, the comfort model is constructed; the specific process is as follows: Obtain the second effective device parameter information set and the physiological parameter information set: associate the rows or columns of the second effective device correlation matrix and the effective physiological parameter correlation matrix respectively through the comfort factor to obtain the second effective device parameter information set and the physiological parameter information set belonging to the comfort factor. For example, for the "low comfort" factor, the seat height, tilt angle, heart rate, blood pressure and other parameter information with high correlation can be obtained. Construct a comfort model: Construct a comfort model based on the second effective device parameter information set and the physiological parameter information set combined with user behavior characteristics. These parameter information can be used as input features, and the comfort factor can be used as the output label. The comfort model can be obtained by training using a machine learning algorithm (such as random forest, neural network, etc.). The model can receive device parameter information, physiological parameter information and user behavior characteristics as input in real time, and output the current comfort factor. In summary, by calculating the correlation matrix of device parameter information and physiological parameter information relative to the comfort factor and obtaining the effective parameter information set, we can construct a comfort model that comprehensively considers device parameters, physiological parameters and user behavior characteristics. This model can help us understand the comfort status in real time and make adjustments and optimizations based on actual conditions.

[0085] In this scheme, since traditional comfort evaluation often only focuses on a single physiological index or environmental parameter, such as temperature and humidity, and ignores the impact of the user's physiological and behavioral changes on comfort on a time scale, this scheme calculates the comfort value by comprehensively considering the changes in the user's physiological indexes and behavioral characteristics during the project execution, and obtains the comfort factor by binning. This makes the comfort evaluation more comprehensive and multi-dimensional, and can better reflect the user's true feelings; among a large number of equipment parameters and physiological parameters, many parameters have no direct correlation with comfort or have a very low correlation. These redundant parameters not only increase the complexity of data processing, but may also affect the accuracy of the comfort model; this scheme effectively screens out the parameter set that has a significant impact on comfort by calculating the correlation between equipment parameters and physiological parameters and comfort factors, and eliminating elements with correlation values ​​less than the set threshold. This reduces the complexity of data processing and improves the accuracy of the comfort model. Traditional comfort models often only consider static parameter relationships, while ignoring the dynamic changes in the correlation between parameters and comfort; this solution introduces a time scale in the correlation calculation, and dynamically considers the changes in the correlation between parameters and comfort over time. This makes the comfort model more consistent with the dynamic changes in actual scenarios, and improves the accuracy and adaptability of the model. Due to the large number of parameters and the complexity of correlation calculation, it is a technical challenge to effectively extract parameter sets that have a significant impact on comfort; this solution successfully extracts effective equipment parameter information sets and physiological parameter information sets belonging to different comfort factors by associating comfort factors with effective equipment parameter correlation matrices and effective physiological parameter correlation matrices, providing strong support for building a more accurate comfort model; by overcoming technical problems such as the lack of multi-dimensional considerations in comfort evaluation, low correlation between parameter redundancy and comfort, lack of dynamics in comfort models, and difficulty in extracting effective parameter sets, a more comprehensive, accurate and dynamic comfort evaluation is achieved in actual scenarios.

[0086] As a preferred embodiment of the present application, the physiological indicators include skin temperature, heart rate, blood pressure and sweating amount; the behavioral characteristic changes include posture adjustment degree, activity level change, and facial expression tension;

[0087] The comfort value is obtained by weighted summation based on skin temperature, heart rate, blood pressure, sweating amount, posture adjustment degree, activity level change, facial expression tension and corresponding weight coefficients.

[0088] In this embodiment, in order to calculate the comfort value, a weighted summation method can be used to multiply the changes in each physiological indicator and behavioral characteristic by their corresponding weight coefficients, and then add the results. The calculation of the comfort value is illustrated by the following specific mathematical formula. The digital features of the following examples cannot be used as a limitation on the actual implementation process and scenario of this solution: comfort value = (skin temperature × w1) + (heart rate × w2) + (blood pressure × w3) + (sweating amount × w4) + (posture adjustment degree × w5) + (activity level change × w6) + (facial expression tension degree × w7); where w1, w2, w3, w4, w5, w6, and w7 are the weight coefficients corresponding to skin temperature, heart rate, blood pressure, sweating amount, posture adjustment degree, activity level change, and facial expression tension degree, respectively. These weight coefficients can be adjusted according to actual conditions to reflect the importance of different indicators to the comfort value.

[0089] Further, a specific example is: assuming the following weight coefficients: skin temperature: w1 = 0.1; heart rate: w2 = 0.2; blood pressure: w3 = 0.2; sweating: w4 = 0.15; posture adjustment degree: w5 = 0.1; activity level change: w6 = 0.1; facial expression tension: w7 = 0.15; and there are the following physiological indicators and behavioral characteristics change measurement values: skin temperature: 30°C; heart rate: 70 times / minute; blood pressure: 120 / 80 mmHg (here the average value of 100 can be taken as a representative); sweating: 0.5 g / min; posture adjustment degree: 2 (assuming a score of 1-5, 5 represents the maximum adjustment); activity level change: 1 (assuming a score of 1-5, 5 represents the maximum activity level change); facial expression tension: 3 (assuming a score of 1-5, 5 represents the most tense); then the comfort value can be obtained by the following calculation: comfort value = (30 × 0.1) + (70 × 0.2) + (100 × 0.2) + (0.5 × 0.15) + (2 × 0.1) + (1 × 0.1) +(3 × 0.15) = 3 + 14 + 20 + 0.075 + 0.2 + 0.1 + 0.45 = 37.825; The comfort value calculated by the example is a relative value, which can be interpreted and applied according to the actual situation; for example, different comfort thresholds can be set to determine whether the current comfort level is high, medium or low. This embodiment obtains a real-time comfort value by collecting the changes in the user's physiological indicators and behavioral characteristics in real time, and performing dynamic weighted summation. This makes the comfort evaluation more in line with the dynamic changes in the actual scene, and can more accurately and comprehensively reflect the user's comfort feeling at different time points.

[0090] As a preferred solution of the present application, a change comfort model is constructed based on the second effective device parameter information set and the physiological parameter information set combined with user behavior characteristics; the steps include:

[0091] Using the second valid device parameter information set and the physiological parameter information set as input data of the fourth machine learning model, using the corresponding comfort factor as output data of the fourth machine learning model, and sequentially training the fourth machine learning model to obtain a fourth weight coefficient value; constructing a third input parameter according to the fourth weight coefficient value, the second valid device parameter information set, and the physiological parameter information set;

[0092] Obtaining user behavior characteristics and a third weight coefficient value on the same time scale as input data of a fifth machine learning model; using the corresponding comfort factor as output data of the fifth machine learning model, and sequentially training the fifth machine learning model to obtain a fifth weight coefficient value; constructing a fourth input parameter according to the fifth weight coefficient value, the user behavior characteristics, and the third input parameter;

[0093] Use the elements in the fourth input parameter as the input of the sixth machine learning model, and use the comfort factor as the output data of the sixth machine learning model to construct an office efficiency model, and obtain the sixth weight coefficient value by training the office efficiency model.

[0094] In this embodiment, first, a variable comfort model is constructed; the specific process is: prepare data for the fourth machine learning model: input data: the second effective device parameter information set (such as seat height, tilt angle, etc.), physiological parameter information set (such as skin temperature, heart rate, etc.). Output data: corresponding comfort factors (such as "low comfort", "medium comfort", "high comfort"). Train the fourth machine learning model: use the prepared input data and output data to train the fourth machine learning model to obtain a fourth weight coefficient value. These weight coefficient values ​​represent the degree of influence of different parameters on the comfort factor. Construct the third input parameter: construct the third input parameter based on the fourth weight coefficient value, the second effective device parameter information set, and the physiological parameter information set. This is a parameter set that comprehensively considers device parameters, physiological parameters and their degree of influence on comfort.

[0095] Secondly, in combination with the user behavior characteristics, prepare the data of the fifth machine learning model: input data: user behavior characteristics on the same time scale (such as the degree of posture adjustment, the change in activity level, etc.), the third weight coefficient value (that is, the weight coefficient value obtained by training the fourth machine learning model). Output data: the corresponding comfort factor. Train the fifth machine learning model: Use the prepared input data and output data to train the fifth machine learning model to obtain the fifth weight coefficient value. These weight coefficient values ​​represent the degree of influence of the user behavior characteristics and the third input parameter on the comfort factor. Construct the fourth input parameter: Construct the fourth input parameter based on the fifth weight coefficient value, user behavior characteristics, and the third input parameter. This is a parameter set that comprehensively considers device parameters, physiological parameters, user behavior characteristics, and their degree of influence on comfort.

[0096] Finally, build the office efficiency model and prepare the data for the sixth machine learning model: Input data: elements in the fourth input parameter. Output data: comfort factor. Train the sixth machine learning model and build the office efficiency model: Use the prepared input data and output data to train the sixth machine learning model, and the obtained sixth weight coefficient value constitutes the office efficiency model. This model can predict the comfort factor based on the input device parameters, physiological parameters and user behavior characteristics, and then evaluate office efficiency.

[0097] A specific example is as follows: Assume that the second valid device parameter information set is: seat height (0.5m), tilt angle (30°). Physiological parameter information set: skin temperature (30°C), heart rate (70 times / minute). User behavior characteristics: posture adjustment degree (2), activity level change (1). Corresponding comfort factor: "medium comfort". The substitution process is as follows: seat height, tilt angle, skin temperature, and heart rate are used as input data of the fourth machine learning model, and "medium comfort" is used as output data for training to obtain a fourth weight coefficient value. The third input parameter is constructed based on the fourth weight coefficient value, seat height, tilt angle, skin temperature, and heart rate. The posture adjustment degree, activity level change, and the third weight coefficient value are used as input data of the fifth machine learning model, and "medium comfort" is used as output data for training to obtain a fifth weight coefficient value. The fourth input parameter is constructed based on the fifth weight coefficient value, posture adjustment degree, activity level change, and the third input parameter. The elements in the fourth input parameter are used as inputs of the sixth machine learning model, and "medium comfort" is used as output data for training to obtain an office efficiency model and a sixth weight coefficient value. Through the above steps, we can build an office efficiency model that comprehensively considers equipment parameters, physiological parameters and user behavior characteristics to evaluate office efficiency under different conditions; so that the variable comfort model can be fully trained and the prediction accuracy and stability of the model can be improved.

[0098] S3, building an office behavior adaptation model based on the office efficiency model and the comfort model; identifying real-time behavior characteristics and combining them with the office behavior adaptation model to output device adjustment strategies, and controlling the smart furniture to perform corresponding actions based on the device adjustment strategies. Specifically, the following steps are included:

[0099] The office efficiency factor, the comfort factor, the third weight coefficient value, the sixth weight coefficient value, the physiological parameter information, and the user behavior characteristics are used as inputs of the seventh machine learning model; the equipment parameter information and the environmental parameter information are used as outputs of the seventh machine learning model to construct an office behavior adaptation model;

[0100] Acquire real-time user behavior feature data, and lock the office efficiency factor with the highest office efficiency and its corresponding third weight coefficient value, the comfort factor with the highest comfort value and its corresponding sixth weight coefficient value, and obtain target device parameter information and target environment parameter information in combination with the office behavior adaptation model;

[0101] A device adjustment strategy is formulated according to target device parameter information and target environment parameter information, and the smart furniture is controlled to perform corresponding actions based on the device adjustment strategy.

[0102] In this embodiment, first, an office behavior adaptation model is constructed; the specific process is: prepare data for the seventh machine learning model: input data: office efficiency factor (such as "high office efficiency", "medium office efficiency", etc.), comfort factor (such as "high comfort", "medium comfort", etc.), third weight coefficient value (from the previous comfort model), sixth weight coefficient value (from the office efficiency model), physiological parameter information (such as skin temperature, heart rate, etc.), user behavior characteristics (such as posture adjustment degree, activity level change, etc.). Output data: equipment parameter information (such as seat height, tilt angle, etc.), environmental parameter information (such as indoor temperature, light intensity, etc.). Train the seventh machine learning model: Use the prepared input data and output data to train the seventh machine learning model to construct an office behavior adaptation model. This model can output corresponding equipment parameter information and environmental parameter information based on the input office efficiency factor, comfort factor, weight coefficient value, physiological parameter information and user behavior characteristics.

[0103] Secondly, identify real-time behavior characteristics and obtain target parameter information; the specific process is: Obtain real-time user behavior characteristic data: collect user behavior characteristic data such as the degree of posture adjustment and activity level change in real time. Lock the optimal office efficiency factor and comfort factor: Based on the real-time user behavior characteristic data, combined with the office efficiency model and the comfort model, lock the office efficiency factor with the highest office efficiency and its corresponding third weight coefficient value, as well as the comfort factor with the highest comfort value and its corresponding sixth weight coefficient value. Obtain target parameter information: Substitute the locked office efficiency factor, comfort factor, weight coefficient value and real-time user behavior characteristic data into the office behavior adaptation model to obtain the target device parameter information and target environment parameter information.

[0104] Finally, formulate device adjustment strategies and control smart furniture; formulate device adjustment strategies: formulate specific device adjustment strategies based on target device parameter information and target environment parameter information. For example, if the target seat height is 0.55m, then the device adjustment strategy is "adjust the seat height to 0.55m". Control smart furniture to perform corresponding actions: send the device adjustment strategy to the smart furniture system to control smart furniture to perform corresponding actions. For example, the smart seat will automatically adjust to a height of 0.55m according to the device adjustment strategy.

[0105] A specific example is as follows: Assume real-time user behavior feature data: posture adjustment degree (3), activity level change (2). The maximum office efficiency factor output by the office efficiency model is: "high office efficiency" and its corresponding third weight coefficient value (0.6). The maximum comfort factor output by the comfort model is: "high comfort" and its corresponding sixth weight coefficient value (0.7). Physiological parameter information: skin temperature (31°C), heart rate (72 times / minute). The substitution process is as follows: "high office efficiency", "high comfort", 0.6, 0.7, skin temperature (31°C), heart rate (72 times / minute), posture adjustment degree (3), activity level change (2) are used as inputs of the seventh machine learning model. The seventh machine learning model outputs target device parameter information and target environment parameter information, such as seat height (0.55m), tilt angle (35°), indoor temperature (24°C), and light intensity (500lx). Formulate a device adjustment strategy: "Adjust the seat height to 0.55m, the tilt angle to 35°, the indoor temperature to 24°C, and the light intensity to 500lx". Send the device adjustment strategy to the smart furniture system to control the smart furniture to perform corresponding actions. This embodiment identifies the real-time behavior characteristics of the user and combines the output of the office behavior adaptation model to formulate a personalized device adjustment strategy to control the smart furniture to perform corresponding actions, making the control of smart furniture more intelligent and personalized, and able to meet the real-time needs and comfort of the user.

[0106] Embodiment 2

[0107] A technical solution also provided in an embodiment of the present invention is a smart furniture comfort adjustment system, comprising:

[0108] A first calculation module: obtaining a first correlation matrix of device parameter information and environmental parameter information relative to an office efficiency factor, and obtaining a first valid device parameter information set and an environmental parameter information set according to the first correlation matrix;

[0109] The first construction module: constructs an office efficiency model based on the first valid device parameter information set and the environment parameter information set combined with user behavior characteristics;

[0110] The second calculation module is used to obtain a second correlation matrix of the device parameter information and the physiological parameter information relative to the comfort factor, and obtain a second valid device parameter information set and a physiological parameter information set according to the second correlation matrix;

[0111] The second building module: building a comfort model based on the second effective device parameter information set and the physiological parameter information set combined with user behavior characteristics;

[0112] The second building module: building an office behavior adaptation model based on the office efficiency model and the comfort model;

[0113] Execution model: Identify real-time behavior characteristics and combine them with the office behavior adaptation model to output device adjustment strategies, and control smart furniture to perform corresponding actions based on the device adjustment strategies.

[0114] In this embodiment, by constructing a correlation matrix, the present application can effectively screen out equipment parameters, environmental parameters and physiological parameters that have a significant impact on office efficiency and comfort, reduce the data processing burden, and improve system efficiency; incorporate user behavior characteristics so that the constructed office efficiency model and comfort model are closer to the actual needs of users, thereby generating more personalized equipment adjustment strategies and improving user experience; by constructing an office behavior adaptation model, the two dimensions of office efficiency and comfort are comprehensively considered to achieve comprehensive optimization of the smart furniture system, which not only improves work efficiency but also ensures user comfort; real-time identification of user behavior characteristics, combined with the office behavior adaptation model, and dynamic output of equipment adjustment strategies, so that smart furniture can respond to user needs immediately and create a more suitable office environment.

[0115] Embodiment 3

[0116] A technical solution provided in an embodiment of the present invention is an electronic device, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor calls the computer program in the memory, the steps of the smart furniture comfort adjustment method based on behavior feature recognition are implemented.

[0117] Embodiment 4

[0118] A technical solution provided in an embodiment of the present invention is a storage medium, in which computer executable instructions are stored. When the computer executable instructions are loaded and executed by a processor, the steps of the smart furniture comfort adjustment method based on behavior feature recognition are implemented.

[0119] Through the description of the above implementation methods, technical personnel in the relevant field can understand that for the convenience and simplicity of description, only the division of the above-mentioned functional modules is used as an example. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the specific device can be divided into different functional modules to complete all or part of the functions described above.

[0120] In the embodiments provided in the present application, it should be understood that the disclosed structures and methods can be implemented in other ways. For example, the embodiments of the structure described above are only schematic. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another structure, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, structures or units, which can be electrical, mechanical or other forms.

[0121] The units described as separate components may or may not be physically separated, and the components shown as units may be one physical unit or multiple physical units, that is, they may be located in one place or distributed in multiple different places. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0122] In addition, each functional unit in the embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0123] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a readable storage medium. Based on this understanding, the technical solution of the embodiment of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium, including several instructions to enable a device (which can be a single-chip microcomputer, chip, etc.) or a processor (processor) to perform all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.

[0124] The specific implementation described above is a preferred implementation of the intelligent furniture comfort adjustment method and system based on behavioral feature recognition of the present invention, and is not intended to limit the specific implementation scope of the present invention. The scope of the present invention includes but is not limited to this specific implementation. All equivalent changes made in accordance with the shape and structure of the present invention are within the protection scope of the present invention.

Claims

1. A method for adjusting the comfort level of intelligent furniture based on behavioral feature recognition, characterized in that: The steps include: Obtaining a first correlation matrix of device parameter information and environmental parameter information relative to an office efficiency factor, and obtaining a first valid device parameter information set and an environmental parameter information set according to the first correlation matrix; Building an office efficiency model based on the first valid device parameter information set and the environment parameter information set combined with user behavior characteristics; Obtain a second correlation matrix of device parameter information and physiological parameter information relative to the comfort factor, and obtain a second valid device parameter information set and physiological parameter information set according to the second correlation matrix; Building a comfort model based on the second effective device parameter information set and the physiological parameter information set combined with user behavior characteristics; Build an office behavior adaptation model based on the office efficiency model and comfort model; identify real-time behavior features and output device adjustment strategies based on the office behavior adaptation model, and control smart furniture to perform corresponding actions based on the device adjustment strategies; The first correlation matrix of the device parameter information and the environmental parameter information relative to the office efficiency factor is obtained, and a first valid device parameter information set and an environmental parameter information set are obtained according to the first correlation matrix; The steps include: The office efficiency value is calculated based on the efficiency ratio, quality coefficient, and completion degree of the project on a time scale, and the office efficiency value is binned to obtain several office efficiency intervals, and each office efficiency interval is used as an office efficiency factor; Extract the equipment parameter information and environmental parameter information corresponding to each office efficiency factor on a time scale; On a time scale, the correlation between the equipment parameter information and the office efficiency factor is calculated according to the correlation algorithm to obtain a first equipment parameter correlation matrix; the elements with correlation values ​​less than a set threshold a1 in the first equipment parameter correlation matrix are removed to obtain a first effective equipment correlation matrix; Synchronously, the correlation between the environmental parameter information and the office efficiency factor is calculated on a time scale according to a correlation algorithm to obtain an environmental parameter correlation matrix; the elements in the environmental parameter correlation matrix whose correlation values ​​are less than a set threshold a2 are removed to obtain an effective environmental parameter correlation matrix; wherein the first correlation matrix includes a first effective device correlation matrix and an effective environmental parameter correlation matrix; By associating the rows or columns of the first effective device correlation matrix and the effective environment parameter correlation matrix respectively through the office efficiency factor, a first effective device parameter information set and an environment parameter information set belonging to the office efficiency factor are obtained; Obtaining a second correlation matrix of device parameter information and physiological parameter information relative to the comfort factor, and obtaining a second valid device parameter information set and a physiological parameter information set according to the second correlation matrix; comprising the following steps: The comfort value is calculated according to the changes in physiological indicators and behavioral characteristics during the execution of the project on a time scale, and the comfort value is binned to obtain several comfort intervals, and each comfort interval is used as a comfort factor; Extract the equipment parameter information and environmental parameter information corresponding to each comfort factor on a time scale; On a time scale, the correlation between the equipment parameter information and the comfort factor is calculated according to the correlation algorithm to obtain a second equipment parameter correlation matrix; the elements with correlation values ​​less than a set threshold a3 in the second equipment parameter correlation matrix are removed to obtain a second effective equipment correlation matrix; Synchronously, the correlation between the physiological parameter information and the comfort factor is calculated on a time scale according to a correlation algorithm to obtain a physiological parameter correlation matrix; the elements in the physiological parameter correlation matrix whose correlation values ​​are less than a set threshold a4 are removed to obtain an effective physiological parameter correlation matrix; wherein the second correlation matrix includes a second effective device correlation matrix and an effective physiological parameter correlation matrix; The rows or columns of the second effective device correlation matrix and the effective physiological parameter correlation matrix are respectively associated with the comfort factor to obtain the second effective device parameter information set and the physiological parameter information set belonging to the comfort factor.

2. The method for adjusting the comfort level of intelligent furniture based on behavioral feature recognition according to claim 1 is characterized in that: The method of constructing an office efficiency model based on the first valid device parameter information set and the environment parameter information set in combination with user behavior characteristics comprises the following steps: Using the first valid device parameter information set and the environmental parameter information set as input data of the first machine learning model, using the corresponding office efficiency factor as output data of the first machine learning model, and sequentially training the first machine learning model to obtain a first weight coefficient value; constructing a first input parameter based on the first weight coefficient value, the first valid device parameter information set, and the environmental parameter information set; Obtaining user behavior characteristics and a first weight coefficient value on the same time scale as input data of a second machine learning model; using a corresponding office efficiency factor as output data of the second machine learning model, and sequentially training the second machine learning model to obtain a second weight coefficient value; constructing a second input parameter according to the second weight coefficient value, the user behavior characteristics, and the first input parameter; The elements in the second input parameter are used as the input of the third machine learning model, and the office efficiency factor is used as the output data of the third machine learning model to construct an office efficiency model, and the third weight coefficient value is obtained by training the office efficiency model.

3. The method for adjusting the comfort level of intelligent furniture based on behavioral feature recognition according to claim 1, characterized in that: Physiological indicators include skin temperature, heart rate, blood pressure and sweating amount; the behavioral characteristic changes include posture adjustment degree, activity level change, and facial expression tension; The comfort value is obtained by weighted summation based on skin temperature, heart rate, blood pressure, sweating amount, posture adjustment degree, activity level change, facial expression tension and corresponding weight coefficients.

4. The method for adjusting the comfort level of intelligent furniture based on behavioral feature recognition according to claim 3 is characterized in that: Constructing a change comfort model based on the second effective device parameter information set and the physiological parameter information set combined with user behavior characteristics; comprising the following steps: Using the second valid device parameter information set and the physiological parameter information set as input data of the fourth machine learning model, using the corresponding comfort factor as output data of the fourth machine learning model, and sequentially training the fourth machine learning model to obtain a fourth weight coefficient value; constructing a third input parameter according to the fourth weight coefficient value, the second valid device parameter information set, and the physiological parameter information set; Obtaining user behavior characteristics and a third weight coefficient value on the same time scale as input data of a fifth machine learning model; using the corresponding comfort factor as output data of the fifth machine learning model, and sequentially training the fifth machine learning model to obtain a fifth weight coefficient value; constructing a fourth input parameter according to the fifth weight coefficient value, the user behavior characteristics, and the third input parameter; Use the elements in the fourth input parameter as the input of the sixth machine learning model, and use the comfort factor as the output data of the sixth machine learning model to construct an office efficiency model, and obtain the sixth weight coefficient value by training the office efficiency model.

5. The method for adjusting the comfort level of intelligent furniture based on behavioral feature recognition according to claim 3 is characterized in that: An office behavior adaptation model is constructed based on the office efficiency model and the comfort model; the real-time behavior characteristics are identified and combined with the office behavior adaptation model to output the device adjustment strategy, and the smart furniture is controlled to perform corresponding actions based on the device adjustment strategy; the steps include: The office efficiency factor, the comfort factor, the third weight coefficient value, the sixth weight coefficient value, the physiological parameter information, and the user behavior characteristics are used as inputs of the seventh machine learning model; the equipment parameter information and the environmental parameter information are used as outputs of the seventh machine learning model to construct an office behavior adaptation model; Acquire real-time user behavior feature data, and lock the office efficiency factor with the highest office efficiency and its corresponding third weight coefficient value, the comfort factor with the highest comfort value and its corresponding sixth weight coefficient value, and obtain target device parameter information and target environment parameter information in combination with the office behavior adaptation model; A device adjustment strategy is formulated according to target device parameter information and target environment parameter information, and the smart furniture is controlled to perform corresponding actions based on the device adjustment strategy.

6. A smart furniture comfort adjustment system, applicable to the smart furniture comfort adjustment method based on behavior feature recognition as described in any one of claims 1 to 5, characterized in that: include: A first calculation module: obtaining a first correlation matrix of device parameter information and environmental parameter information relative to an office efficiency factor, and obtaining a first valid device parameter information set and an environmental parameter information set according to the first correlation matrix; The first construction module: constructs an office efficiency model based on the first valid device parameter information set and the environment parameter information set combined with user behavior characteristics; The second calculation module is used to obtain a second correlation matrix of the device parameter information and the physiological parameter information relative to the comfort factor, and obtain a second valid device parameter information set and a physiological parameter information set according to the second correlation matrix; The second building module: building a comfort model based on the second effective device parameter information set and the physiological parameter information set combined with user behavior characteristics; The second building module: building an office behavior adaptation model based on the office efficiency model and the comfort model; Execution model: Identify real-time behavior characteristics and combine them with the office behavior adaptation model to output device adjustment strategies, and control smart furniture to perform corresponding actions based on the device adjustment strategies.

7. An electronic device, characterized in that: It comprises a memory and a processor, wherein the memory stores a computer program, and when the processor calls the computer program in the memory, the steps of the intelligent furniture comfort adjustment method based on behavior feature recognition as described in any one of claims 1 to 5 are implemented.

8. A storage medium, characterized in that: The storage medium stores computer executable instructions, and when the computer executable instructions are loaded and executed by the processor, the steps of the smart furniture comfort adjustment method based on behavior feature recognition as described in any one of claims 1 to 5 are implemented.

Images