Fluid motion control method and system based on axial flow fan

Abstract

The invention discloses a fluid motion control method and system based on an axial flow fan, and the method specifically comprises the steps: carrying out the processing of fluid state data and human motion data through a space-time alignment algorithm, and generating a joint representation model; based on the joint characterization model, constructing a fluid simulation model by adopting an LBM technology, and generating a first fluid motion control parameter through the fluid simulation model and a preset multi-objective optimization function; correcting the first fluid motion control parameter through the PMV-PPD thermal comfort index to form a second fluid motion control parameter; and correcting the first fluid motion control parameter or the second fluid motion control parameter through the user emotional state data to form a third fluid motion control parameter. According to the system, high-precision and intelligent control over fluid motion of the axial flow fan is achieved, the sensing capacity and control precision of the system can be improved, and the collaboration, health guarantee capacity and interactive experience of the system can be enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of axial flow fans, and in particular to a fluid motion control method and system based on an axial flow fan. Background Art

[0002] In smart home applications, axial fans, with their high efficiency and low noise, have become essential indoor environmental regulators, widely used in various residential spaces. Despite this, axial fans still face a number of challenges in fluid motion control, hindering their full performance and increasing their intelligence.

[0003] The primary issue lies in the limitations of sensing capabilities. In smart home environments, the fixed layout of traditional sensor networks struggles to adapt to diverse indoor layouts and dynamically changing spatial requirements. This leads to frequent blind spots and an inability to fully and accurately capture indoor fluid conditions (such as temperature, humidity, and wind speed) as well as human activity data. Furthermore, traditional sensor networks lack data integration capabilities, making it difficult to effectively integrate information from diverse physical fields. This limits the ability to deeply understand and precisely control the indoor environment. This situation not only impairs the adaptability of axial fans to complex indoor environments but also reduces the accuracy and flexibility of their intelligent control.

[0004] Secondly, insufficient control precision is another major concern. In smart home scenarios, fluid motion is a complex system involving the interaction of multiple variables. Traditional control strategies often struggle to precisely regulate this dynamic process. This not only leads to high energy consumption during axial flow fan operation, but can also affect occupant comfort due to improper regulation of parameters such as temperature, humidity, and wind speed. Users often experience discomfort due to unbalanced environmental parameters, which in turn raises concerns about the performance and practicality of axial flow fans.

[0005] Furthermore, a lack of system synergy is a common problem in axial fan control systems within smart home environments. In a smart home ecosystem, multiple devices coexist, and the lack of effective coordination between axial fans and other devices can easily lead to mutual interference, impacting overall control effectiveness. This not only limits the full performance of axial fans but also makes it difficult to meet the personalized needs of users in complex environments, reducing the collaborative efficiency and overall effectiveness of the entire smart home system.

[0006] Furthermore, insufficient health protection capabilities are another pressing issue that needs to be addressed. In smart home applications, traditional axial fan control systems lack in-depth modeling and analysis of user health and pathogen transmission pathways, making it difficult to achieve health-oriented fluid control.

[0007] Finally, the limited interactive experience has become a bottleneck hindering the intelligent development of axial flow fans. In smart home scenarios, users expect to interact with devices in a more natural and intelligent way. However, traditional control methods rely too heavily on fixed user interfaces, lacking flexibility and intelligence, resulting in a poor user experience. Users need to manually adjust various axial flow fan parameters, which is cumbersome and difficult to achieve a personalized comfort experience. This has, to a certain extent, hindered the widespread application and further development of axial flow fans in the smart home sector. Summary of the Invention

[0008] The purpose of the present invention is to provide a fluid motion control method and system based on an axial flow fan, which realizes high-precision and intelligent control of the fluid motion of the axial flow fan, which can not only improve the system's perception ability and control accuracy, but also enhance the system's coordination, health protection ability and interactive experience, thereby meeting the user's personalized needs in complex environments and solving at least one of the above-mentioned existing technical problems.

[0009] In a first aspect, the present invention provides a method for controlling fluid motion based on an axial flow fan, the method specifically comprising:

[0010] A sensor network is used to obtain indoor fluid state data and human motion data in real time. The fluid state data and human motion data are processed through a spatiotemporal alignment algorithm to generate a joint representation model.

[0011] Based on the joint characterization model, a fluid simulation model is constructed using LBM technology, and the first fluid motion control parameters of the axial flow fan are generated through the fluid simulation model and a preset multi-objective optimization function;

[0012] A personalized metabolic rate prediction model is constructed based on the human metabolic data of indoor occupants, a PMV-PPD thermal comfort index is calculated in real time using the personalized metabolic rate prediction model, and the first fluid motion control parameter is corrected using the PMV-PPD thermal comfort index to form a second fluid motion control parameter for the axial flow fan;

[0013] Multimodal emotion recognition technology is used to analyze the user emotional state data of indoor occupants, and the first fluid motion control parameter or the second fluid motion control parameter is corrected based on the user emotional state data to form the third fluid motion control parameter of the axial flow fan.

[0014] In a second aspect, the present invention provides a fluid motion control system based on an axial flow fan, the system specifically comprising:

[0015] The first control module is used to use a sensor network to obtain indoor fluid state data and human motion data in real time, process the fluid state data and human motion data through a spatiotemporal alignment algorithm, and generate a joint representation model;

[0016] A second control module is configured to construct a fluid simulation model based on the joint characterization model using LBM technology, and generate first fluid motion control parameters of the axial flow fan using the fluid simulation model and a preset multi-objective optimization function;

[0017] a third control module, configured to construct a personalized metabolic rate prediction model based on the human metabolic data of the indoor occupants, calculate the PMV-PPD thermal comfort index in real time using the personalized metabolic rate prediction model, and correct the first fluid motion control parameter using the PMV-PPD thermal comfort index to form a second fluid motion control parameter for the axial flow fan;

[0018] The fourth control module is used to use multimodal emotion recognition technology to analyze the user emotional state data of indoor people, and to correct the first fluid motion control parameter or the second fluid motion control parameter based on the user emotional state data to form the third fluid motion control parameter of the axial flow fan.

[0019] In a third aspect, the present invention provides a computer device comprising: a memory and a processor and a computer program stored in the memory, wherein when the computer program is executed on the processor, the fluid motion control method based on the axial flow fan as described in any one of the above methods is implemented.

[0020] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the fluid motion control method based on an axial flow fan as described in any one of the above methods is implemented.

[0021] Compared with the prior art, the present invention has at least one of the following technical effects:

[0022] 1. The present invention realizes high-precision and intelligent control of the fluid movement of the axial flow fan, which not only improves the system's perception ability and control accuracy, but also enhances the system's coordination, health protection capabilities and interactive experience, thereby meeting the user's personalized needs in complex environments.

[0023] 2. The present invention realizes comprehensive and real-time perception of indoor fluid status and human movement by integrating sensor networks and spatiotemporal alignment algorithms, providing a data basis for precise control.

[0024] 3. The present invention uses LBM technology to construct a fluid simulation model, and combines it with a multi-objective optimization function to generate fluid motion control parameters that are efficient and meet the requirements of the indoor environment.

[0025] 4. The present invention combines the personalized metabolic rate prediction model and the PMV-PPD thermal comfort index to achieve refined adjustment of the user's thermal comfort and improve the user experience.

[0026] 5. The present invention introduces multimodal emotion recognition technology to dynamically adjust control parameters according to the user's emotional state, thereby enhancing the intelligence and personalization level of the system.

[0027] 6. The present invention improves the positioning accuracy and dynamic adaptability of the sensor network through SLAM technology and extended Kalman filtering, ensuring the accuracy and real-time performance of the data.

[0028] 7. The present invention enhances the robustness and accuracy of the joint representation model and improves the accuracy of data fusion by applying the adaptive weight matrix and Gaussian kernel density estimation method.

[0029] 8. The present invention makes the fluid simulation model more accurate and efficient through the application of LBM technology and the setting of boundary conditions, and can truly reflect the indoor fluid movement.

[0030] 9. The present invention ensures that the initial state of the simulation conforms to the actual situation through the initialization method of the distribution function, thereby improving the authenticity and accuracy of the simulation.

[0031] 10. The present invention combines a multi-objective optimization function with a parameter adaptive calibration function to achieve comprehensive optimization of the axial flow fan control parameters, thereby improving the overall performance and stability of the system.

[0032] 11. The present invention achieves accurate calculation of user thermal comfort and improves user experience through the application of personalized metabolic rate prediction model and PMV-PPD calculation model.

[0033] 12. The present invention dynamically adjusts the control parameters according to the PMV-PPD thermal comfort index, thereby improving the system's response speed and comfort control capability.

[0034] 13. The present invention combines multimodal emotion recognition technology and the second control parameter correction function to achieve real-time monitoring and personalized adjustment of the user's emotional state, enhancing the intelligence of the system and user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0036] Figure 1 1 is a flow chart of a method for controlling fluid motion based on an axial flow fan provided by one embodiment of the present invention;

[0037] Figure 2 1 is a schematic structural diagram of a fluid motion control system based on an axial flow fan provided by one embodiment of the present invention;

[0038] Figure 3 It is a structural diagram of a computer device provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0039] In the following description, specific details such as specific system structures and techniques are provided for purposes of illustration rather than limitation to facilitate a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid obscuring the description of the present application with unnecessary detail.

[0040] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.

[0041] It will also be understood that the term "and / or" used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.

[0042] As used in this specification and the appended claims, the term "if" can be interpreted as "when" or "upon" or "in response to determining" or "in response to detecting," depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "upon determination" or "in response to determining" or "upon detection of [described condition or event]" or "in response to detecting [described condition or event]," depending on the context.

[0043] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0044] References to "one embodiment" or "some embodiments" in this specification mean that a particular feature, structure, or characteristic described in conjunction with that embodiment is included in one or more embodiments of the present application. Thus, phrases such as "in one embodiment," "in some embodiments," "in other embodiments," and "in other embodiments" appearing in various places in this specification do not necessarily refer to the same embodiment, but rather mean "one or more but not all embodiments," unless otherwise specifically emphasized. The terms "including," "comprising," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0045] In the embodiments of the present application, the execution subject of the process includes a terminal device, which includes but is not limited to: a server, a computer, a smart phone, a tablet computer, and other devices capable of executing the method disclosed in the present application. Figure 1 A flow chart of a fluid motion control method based on an axial flow fan disclosed in one embodiment of the present invention is shown, and is described in detail as follows:

[0046] S101 uses a sensor network to obtain indoor fluid state data and human motion data in real time, processes the fluid state data and human motion data through a spatiotemporal alignment algorithm, and generates a joint representation model.

[0047] In this embodiment, a sensor network consisting of multiple sensors is deployed in a smart home environment to achieve intelligent control of the indoor environment. These sensors include temperature sensors, humidity sensors, wind speed sensors, and human infrared sensors. The temperature and humidity sensors monitor indoor temperature and humidity in real time, reflecting the basic state of the fluid; the wind speed sensor measures indoor wind speed and provides fluid motion data; and the human infrared sensor detects the movement and position of people indoors, acquiring human motion data.

[0048] In practice, the sensor network collects data at a constant frequency and sends it to a central processing unit (CPU). Upon receiving the data, the CPU first processes the fluid state data and human motion data using a spatiotemporal alignment algorithm. This algorithm primarily involves two steps: time synchronization and spatial calibration. Time synchronization ensures that all data is compared and analyzed on the same time basis by matching timestamps. Spatial calibration utilizes known sensor location information to spatially unify data collected by different sensors into a consistent coordinate system.

[0049] After the spatiotemporal alignment, the central processing unit further generates a joint representation model using a Gaussian mixture model (or similar methods). This model fuses the fluid state data and human motion data into a multidimensional feature vector, with each feature vector corresponding to a grid cell in the indoor space. This joint representation model allows for intuitive visualization of the correlation and distribution between indoor fluid states and human motion.

[0050] In this embodiment, the spatiotemporal alignment algorithm ensures the temporal and spatial consistency of fluid state data and human motion data, providing a reliable foundation for subsequent data analysis and processing. The joint representation model integrates data from different sources into a single entity, enabling the system to more comprehensively understand the indoor environment and occupant activity, thereby making more intelligent control decisions. By monitoring and analyzing indoor fluid state and human motion data in real time, the system can promptly adjust indoor environmental parameters (such as temperature, humidity, and wind speed) to meet user comfort and health needs, thereby enhancing the user experience. The joint representation model can handle multiple data types and sources, enhancing the system's flexibility and robustness, enabling it to adapt to diverse indoor environments and user behavior patterns.

[0051] S102 , based on the joint characterization model, a fluid simulation model is constructed using LBM technology, and first fluid motion control parameters of the axial flow fan are generated through the fluid simulation model and a preset multi-objective optimization function.

[0052] In this example, the grid cells in the joint representation model are first mapped onto the LBM computational network. Each grid cell corresponds to a node in the LBM network, and the node's distribution function describes the fluid's motion in different directions. Based on the data in the joint representation model, these distribution functions are initialized to reflect the initial fluid state in the room.

[0053] Next, set the axial fan's outlet velocity boundary condition and dynamic obstacle boundary condition. The outlet velocity boundary condition is calculated based on parameters such as the fan's speed and blade diameter and is used to simulate the airflow generated by the axial fan. The dynamic obstacle boundary condition is set based on obstacle information (such as human bodies and furniture) in the joint representation model to simulate the impact of these obstacles on fluid motion.

[0054] During the LBM simulation process, the distribution function of each node is continuously updated, simulating the motion of the fluid through collision and migration steps. The collision step simulates the interaction between fluid particles, while the migration step propagates the motion state of fluid particles to adjacent nodes. Through continuous iterative calculations, the dynamic distribution of the fluid in the room can be determined.

[0055] To generate the primary fluid motion control parameters for the axial fan, a multi-objective optimization function is pre-set. These functions include comfort objectives (such as optimizing parameters like temperature, humidity, and wind speed), energy consumption objectives (such as minimizing the fan's energy consumption), and equipment life objectives (such as avoiding prolonged high-speed operation of the fan to reduce wear). Using the fluid dynamic distribution data obtained from the fluid simulation model, combined with these optimization functions, a multi-objective optimization algorithm (such as genetic algorithm, particle swarm optimization, or NSGA-III) is employed to determine the optimal axial fan control parameters that meet all objectives.

[0056] In this embodiment, by combining the joint characterization model and LBM technology, the dynamic distribution of fluid in the room can be simulated more accurately, thereby obtaining more reliable control parameters. The pre-set multi-objective optimization function makes it possible to comprehensively consider multiple aspects such as comfort, energy consumption and equipment life to obtain the optimal axial flow fan control parameters. Due to the use of real-time fluid state data and human motion data, and the ability to dynamically adjust the control parameters of the axial flow fan, the system can better adapt to different indoor environments and user behavior patterns. By optimizing the control parameters of the axial flow fan, a more comfortable indoor environment can be achieved while reducing energy consumption and equipment wear, thereby improving the overall user experience.

[0057] S103, constructing a personalized metabolic rate prediction model based on the human metabolic data of indoor occupants, using the personalized metabolic rate prediction model to calculate the PMV-PPD thermal comfort index in real time, and correcting the first fluid motion control parameter using the PMV-PPD thermal comfort index to form the second fluid motion control parameter of the axial flow fan.

[0058] In this embodiment, in order to improve indoor thermal comfort and energy efficiency in a smart home or office environment, an axial fan control strategy combining a personalized metabolic rate prediction model and the PMV-PPD thermal comfort index is adopted.

[0059] First, a personalized metabolic rate prediction model is constructed by collecting metabolic data from indoor occupants. This data, including heart rate, activity level, weight, height, and other information, can be obtained through wearable devices (such as heart rate monitors and fitness trackers) or professional medical equipment. Based on this data, a personalized metabolic rate prediction model is trained using machine learning algorithms (such as linear regression, decision trees, and random forests). This model accurately predicts the metabolic rate of indoor occupants based on their real-time activity level and physical characteristics.

[0060] Next, the personalized metabolic rate prediction model was combined with the PMV-PPD thermal comfort index. The PMV-PPD index is a standard tool for assessing indoor thermal comfort. PMV stands for predicted mean vote value, reflecting people's thermal perception of the indoor environment; PPD stands for predicted percentage dissatisfaction, indicating the proportion of people dissatisfied with the thermal environment. Based on international standards such as ASHRAE 55 and ISO 7730, the PMV-PPD index is calculated in real time, combining actual indoor environmental parameters (such as temperature, humidity, wind speed, and radiant temperature) with the predicted metabolic rate.

[0061] The calculated PMV-PPD thermal comfort index is then used to modify the primary fluid motion control parameters of the axial fan. If the PMV value deviates from the comfortable range (-0.5 to +0.5) or the PPD value is too high (over 20%), it indicates that the current indoor environment needs adjustment. Based on the PMV-PPD feedback, the axial fan speed, outlet angle, and other parameters are adjusted to improve the indoor temperature and air velocity distribution, thereby enhancing thermal comfort.

[0062] After correction, the second fluid motion control parameters of the axial fan are obtained. These parameters can more accurately meet the thermal comfort needs of indoor occupants while reducing energy consumption and equipment wear.

[0063] In this embodiment, by combining the personalized metabolic rate prediction model and the PMV-PPD indicator, the thermal comfort needs of indoor occupants can be more accurately assessed, and the control parameters of the axial flow fan can be adjusted according to the needs, thereby improving the overall thermal comfort. Since the operation of the axial flow fan can be adjusted according to real-time environmental parameters and occupant activity levels, unnecessary energy consumption can be avoided. When the indoor environment is already comfortable enough, the axial flow fan can reduce its speed or be turned off, thereby saving energy. This embodiment combines a variety of technologies and algorithms, including machine learning, thermal comfort assessment, and intelligent control, so that the system can adapt to different indoor environments and occupant needs more intelligently. By providing personalized thermal comfort and intelligent axial flow fan control strategies, this embodiment can enhance the user's overall experience, allowing them to work and live in a more comfortable environment.

[0064] S104, using multimodal emotion recognition technology to analyze the user emotional state data of indoor occupants, and correcting the first fluid motion control parameter or the second fluid motion control parameter based on the user emotional state data to form a third fluid motion control parameter of the axial flow fan.

[0065] In this embodiment, in an advanced smart home system, in order to improve the comfort of the indoor environment and meet the personalized needs of users, multimodal emotion recognition technology is introduced to analyze the user emotional state data of indoor occupants, and the fluid motion control parameters of the axial fan are corrected accordingly.

[0066] In practice, sensors such as high-definition cameras, microphones, and wearable devices deployed indoors capture multimodal data, including facial expressions, speech, and physiological signals (such as heart rate and respiratory rate), in real time. This data is transmitted to a central processing unit, where deep learning algorithms are used for emotion recognition. This multimodal emotion recognition technology combines face detection with expression recognition, speech emotion analysis, and physiological signal analysis to more accurately understand the user's emotional state.

[0067] Once the user's emotional state (such as joy, sadness, anxiety, etc.) is identified, the first or second fluid motion control parameters of the axial flow fan are modified based on the preset correspondence between emotion and fluid motion control parameters. For example, if the user displays a joyful emotional state, the current fluid motion control parameters may be maintained or fine-tuned to maintain indoor comfort. If the user displays anxiety or uneasiness, the axial flow fan's speed may be increased or the air outlet angle adjusted to improve indoor ventilation and temperature distribution, thereby alleviating the user's discomfort.

[0068] After correction, the third fluid motion control parameters for the axial fan are obtained. These parameters not only take into account the indoor physical environment (such as temperature, humidity, wind speed, etc.), but also incorporate the user's emotional state, making the indoor environment more personalized to the user's needs.

[0069] In this embodiment, by combining multimodal emotion recognition technology and axial flow fan control strategies, it is possible to more accurately understand the user's emotional needs and adjust the indoor environment according to their needs, thereby improving the user's overall satisfaction. This embodiment enables the smart home system to dynamically adjust the indoor environment according to the user's emotional state, enhancing the system's adaptability and flexibility. The application of multimodal emotion recognition technology promotes interaction between people and smart home systems, enabling the system to respond more intelligently to user needs and emotional changes. By optimizing the control parameters of the axial flow fan, the indoor ventilation and temperature distribution are improved, allowing users to live and work in a more comfortable environment and enhance the living experience.

[0070] In some embodiments, in step S101, processing the fluid state data and the human motion data using a spatiotemporal alignment algorithm to generate a joint representation model specifically includes:

[0071] The SLAM technology is used to calibrate the spatial position of each sensor node in the room, and the extended Kalman filter is used to update the position of the self-moving sensor nodes in real time;

[0072] An interpolation algorithm is used to unify the fluid state data and human motion data collected by the sensor network to the same time base;

[0073] The room is divided into multiple grid cells, and the fluid state data and human motion data are mapped to the grid cells through Gaussian kernel density estimation to form a joint representation model.

[0074] In this embodiment, multiple sensor nodes, including fixed nodes and self-moving nodes, are deployed in an indoor environment. SLAM (Simultaneous Localization and Mapping) technology is used for environmental perception and map construction. Either LiDAR SLAM or visual SLAM can be used for this purpose, with the specific choice depending on the complexity and lighting conditions of the indoor environment. For fixed sensor nodes, their precise position in the map can be directly determined using SLAM technology. For self-moving sensor nodes, after the initial position is determined using SLAM technology, the extended Kalman filter (EKF) algorithm is used to update their position in real time. The extended Kalman filter combines the two steps of prediction and update, and can effectively handle nonlinear motion models and sensor noise.

[0075] Extended Kalman filter satisfies , , represents the estimated position of the sensor node at time k, represents the actual position of the sensor node at time k-1, represents the moving speed of the sensor node at time k, Indicates the time interval, represents the position noise, Represents a three-dimensional vector.

[0076] Sensor networks collect fluid state data (such as temperature, humidity, and flow rate) and human motion data (such as position, velocity, and acceleration). Because different sensors may have different sampling frequencies and clock drift, the data may be time-desynchronized. Interpolation algorithms (such as linear or spline interpolation) are used to align fluid state and human motion data to the same time base. Interpolation algorithms estimate the values ​​of unknown data points based on known data points, thereby achieving data synchronization.

[0077] The interpolation algorithm satisfies ,in, represents the time-aligned measurement value, Indicates that the sensor is at time The original measurement value of , t represents the target time point after alignment, Indicates the time point of sensor data collection, Indicates the maximum sampling interval of the sensor, represents the interpolation kernel function.

[0078] The indoor space is divided into multiple grid cells, each representing a spatial region. The fluid state data and human motion data are mapped to the grid cells using Gaussian Kernel Density Estimation (KDE). KDE is a nonparametric method that estimates the probability density function of the data. By calculating the data density within each grid cell, a joint representation model is formed. This model reflects the spatial distribution and interrelationships of the fluid state data and human motion data.

[0079] This implementation improves the accuracy and reliability of indoor environment perception, achieves data synchronization and spatial fusion, and provides comprehensive data support for intelligent decision-making and control systems. The joint representation model helps discover the potential relationships between the indoor environment and human behavior, providing a scientific basis for optimizing indoor environmental design and improving living comfort.

[0080] In some embodiments, the joint representation model satisfies

[0081]

[0082] in, represents the multidimensional feature vector of the grid unit (h, w, d) at time t, K represents the total number of sensors, represents the measurement value of the kth sensor, represents the spatial position of the kth sensor, represents the center coordinates of the grid cell (h,w,d), represents the spatial smoothing coefficient, represents the Euclidean distance between the sensor location and the center of the grid cell;

[0083]

[0084] in, represents the adaptive weight matrix used to adjust the contribution of sensor data to the joint representation model, represents the prediction joint representation model calculated by the weight matrix W, represents the true joint representation model obtained through experiments or high-precision measurements, represents the Frobenius norm of the weight matrix and is used to measure the difference between the predicted joint representation model and the true joint representation model, represents the regularization coefficient used to control the sparsity of the weight matrix, represents the L1 norm of the weight matrix.

[0085] In this embodiment, the multidimensional feature vector The multidimensional feature vector of a grid cell (h, w, d) at time t combines measurements from different sensors and reflects the environmental state (such as temperature, humidity, and occupant activity) of that grid cell at that specific point in time. By constructing this multidimensional feature vector, we can fully capture the complexity and dynamics of the indoor environment, providing a rich information foundation for subsequent data analysis and processing.

[0086] Sensor measurement value represents the measurement value of the kth sensor, which directly reflects the physical quantity monitored by the sensor (such as temperature, humidity, and personnel location). Sensor measurements are the basic data for building joint representation models, and their accuracy and reliability directly determine the effectiveness of the model.

[0087] Sensor spatial location represents the spatial position of the kth sensor and is used to determine the spatial relationship between the sensor and the grid cell. By considering the spatial position of the sensor, the contribution of the sensor data to the grid cell feature vector can be more accurately evaluated, thereby improving the accuracy and robustness of the model.

[0088] Grid cell center coordinates This function is used to calculate the Euclidean distance between the sensor location and the grid cell center. The grid cell center coordinates connect sensor data and the grid cell feature vector. This distance calculation allows for a reasonable assessment of the impact of sensor data on the grid cell feature vector.

[0089] Spatial smoothing coefficient It is used to control the degree of spatial smoothing of sensor data on the grid cell feature vector. By adjusting the spatial smoothing coefficient, it is possible to balance the local accuracy and global consistency of the data, thereby improving the generalization ability and adaptability of the model.

[0090] The adaptive weight matrix W is used to adjust the contribution of sensor data to the joint representation model. It dynamically adjusts weights based on factors such as sensor measurements, spatial location, and grid cell feature vectors. This adaptive weight matrix can rationally allocate weights based on the characteristics of different sensors and the requirements of grid cells, thereby improving the accuracy and flexibility of the model.

[0091] Predictive Joint Representation Model Joint representation model with the truth This is used to measure the model's predictive performance. By comparing the predicted joint representation model with the true joint representation model, the accuracy and reliability of the model can be evaluated, providing a basis for model optimization and improvement.

[0092] The Frobenius norm of the weight matrix measures the difference between the predicted joint representation model and the true joint representation model; the L1 norm controls the sparsity of the weight matrix. By introducing the Frobenius norm and L1 norm, we can control the complexity and sparsity of the weight matrix while maintaining model accuracy, thereby improving the model's interpretability and computational efficiency.

[0093] In some embodiments, in the above step S102, constructing a fluid simulation model based on the joint characterization model using LBM technology specifically includes:

[0094] Map the grid cells of the joint representation model to the LBM computing network, and set the distribution function of each grid node of the LBM computing network ,in, It is used to represent the discrete velocity direction index of the LBM model, X=(x,y,z) is used to represent the grid node coordinates, and t represents the simulation time step;

[0095] Set the axial fan outlet velocity boundary condition and dynamic obstacle boundary conditions ,in, represents the velocity field at the outlet of the axial fan, represents the velocity component of the axial fan outlet along the z-axis, Indicates the empirical coefficient used to convert the axial fan speed into the axial fan outlet speed. Indicates the speed of the axial flow fan, D indicates the diameter of the axial flow fan blade, represents the distribution function value at the boundary of a wall or obstacle, Represents the distribution function value in the direction opposite to i, represents the weight coefficient of the discrete velocity direction, represents the fluid density, The vector representing the direction of the i-th discrete velocity, represents the obstacle surface velocity provided by the joint characterization model, represents the speed of sound;

[0096] The distribution function is initialized, and simulation is performed in combination with the axial flow fan air outlet velocity boundary condition and the dynamic obstacle boundary condition to form a fluid simulation model.

[0097] In this embodiment, the grid cells of the joint representation model are mapped to the computational network of the lattice Boltzmann method (LBM), with each grid cell corresponding to a grid node in the LBM network. This mapping process ensures a smooth transition from the joint representation model to the LBM, enabling the LBM to accurately simulate the environmental conditions described by the joint representation model. Through this mapping, the LBM can utilize the multidimensional feature vectors and sensor data provided by the joint representation model to accurately simulate indoor fluid flow. This facilitates a deeper understanding of the fluid dynamics of indoor environments and provides a scientific basis for optimizing indoor environmental design and control strategies.

[0098] In LBM, a distribution function describes the probability distribution of fluid particles along discrete velocity directions. By setting a distribution function, the motion of a fluid in different velocity directions can be simulated. A well-defined distribution function accurately reflects the microscopic motion characteristics of the fluid, allowing macroscopic fluid flow patterns to be derived through statistical averaging. This helps reveal the flow behavior of fluids in complex indoor environments, providing powerful support for indoor environmental control and optimization.

[0099] Axial fan outlet velocity boundary conditions are used to simulate the effects of axial fans on indoor fluid flow. By setting the velocity field at the axial fan outlet, the impact of the fan's airflow on the indoor environment can be simulated. Introducing axial fan outlet velocity boundary conditions accurately simulates the fan's impact on indoor fluid flow, allowing assessment of the fan's contribution to improving the indoor environment and increasing ventilation efficiency. This is crucial for optimizing fan layout and control strategies.

[0100] Dynamic obstacle boundary conditions are used to simulate the effects of moving obstacles (such as people and equipment) on fluid flow in indoor environments. By setting the distribution function value at the obstacle boundary, you can simulate the obstacles' obstruction and guidance effects on fluid flow. Incorporating dynamic obstacle boundary conditions allows for more realistic simulation of fluid flow in indoor environments. This helps assess the impact of obstacles on indoor ventilation, temperature distribution, and other factors, providing a scientific basis for optimizing indoor environment layout and control strategies.

[0101] The distribution function is initialized, and simulation is performed in conjunction with the axial fan outlet velocity boundary conditions and dynamic obstacle boundary conditions. Through iterative calculations, the distribution function is gradually updated, ultimately resulting in a fluid simulation model. The simulation initialization and simulation process simulates the dynamic changes in indoor fluid flow and reveals the flow behavior of fluids under different conditions. This provides a deeper understanding of the fluid dynamics of indoor environments and provides strong support for optimizing indoor environmental design and control strategies. Simulations can also predict the response of indoor environments under different control strategies, providing a scientific basis for practical applications.

[0102] Furthermore, the initializing the distribution function specifically includes:

[0103] When the distribution function At t=0, we get , , in, represents the distribution function value along the direction of the i-th discrete velocity at position X and time t=0, represents the initial fluid density, represents the initial velocity field, represents the equilibrium distribution function calculated from the initial fluid density and initial velocity field, represents the humidity data of the joint characterization model, Represents the wind speed data of the joint representation model.

[0104] In this embodiment, at time t=0, the distribution function value is represents the distribution of fluid particles along the i-th discrete velocity direction at position X and time t = 0. This initial value is the equilibrium distribution function calculated based on the initial fluid density and initial velocity field. , and may be affected by the humidity data H and wind speed data V used in the joint characterization model (although this influence is not explicitly reflected in the direct expression, it may be indirectly reflected by adjusting the parameters or form of the equilibrium distribution function). Setting a reasonable initial distribution function value is the foundation of the simulation process, as it determines the starting point of the simulation process and the fundamental state of the subsequent fluid flow evolution. By considering the initial fluid density, velocity field, and possibly humidity and wind speed data, the initial state of the fluid in a real indoor environment can be more accurately simulated, thereby improving the accuracy and reliability of the simulation.

[0105] Initial fluid density It is one of the fundamental physical quantities that describes the fluid at the start of a simulation. It determines the spatial distribution density of fluid particles, which in turn influences the fluid's flow behavior and characteristics. Accurately setting the initial fluid density is crucial for simulating fluid flow. It directly influences the distribution and evolution of macroscopic physical quantities such as pressure, velocity, and temperature. By incorporating environmental data (such as humidity and wind speed) provided by the joint characterization model, the initial fluid density can be more appropriately set, leading to a more accurate simulation of fluid flow behavior in a real indoor environment.

[0106] Initial velocity field This describes the velocity distribution of the fluid at the start of the simulation. This distribution can be uniform or non-uniform, depending on the actual situation. The choice of the initial velocity field has a significant impact on the evolution of the fluid flow. A reasonable initial velocity field setting can reflect the initial motion state of the fluid in a real indoor environment. By considering factors such as the axial fan outlet velocity boundary conditions and dynamic obstacle boundary conditions, the initial velocity field can be more accurately set, thereby more realistically simulating the flow behavior of the fluid in a complex indoor environment.

[0107] Equilibrium distribution function The equilibrium distribution function is a key concept in the lattice Boltzmann method (LBM). It describes the probability distribution of fluid particles along discrete velocity directions in the absence of external forces. The equilibrium distribution function is typically related to macroscopic physical quantities such as fluid density, velocity, and temperature. Accurately specifying the equilibrium distribution function is crucial for simulating fluid flow. It determines the distribution of fluid particles along discrete velocity directions, thereby influencing the macroscopic flow behavior of the fluid. By considering factors such as the initial fluid density and initial velocity field, a reasonable equilibrium distribution function can be calculated, thereby more accurately simulating the flow behavior of fluids in real indoor environments.

[0108] In some embodiments, in the above step S102, generating the first fluid motion control parameter of the axial flow fan by using the fluid simulation model and a preset multi-objective optimization function specifically includes:

[0109] A first optimization function is set according to the comfort target, and the first optimization function satisfies ,in, represents the first optimization function value, t represents the simulation time step, T represents the total number of simulation time steps, represents the percentage of unsatisfactory predictions over time step t, represents the velocity gradient changing with time step t, represents the sum of squares of velocity gradients varying with time step t, represents the weight coefficient;

[0110] A second optimization function is set according to the energy consumption target, and the second optimization function satisfies ,in, represents the second optimization function value, RPM(t) represents the axial fan speed that changes with time step t, represents the change in the axial fan blade angle as the time step t changes, Indicates the power factor used to convert the axial fan speed into energy consumption, Indicates the power coefficient used to convert the change in the axial fan blade angle into energy consumption;

[0111] A third optimization function is set according to the equipment life target, and the third optimization function satisfies ,in, represents the third optimization function value, Indicates the maximum allowable speed of the axial fan. represents the material fatigue index, represents the speed variance of the axial fan, represents the stability weight coefficient used to balance the speed and speed variance of the axial fan;

[0112] A parameter adaptive calibration function is set, and the parameter adaptive calibration function satisfies ,in, Indicates the adjusted axial fan speed, represents the optimal speed of the axial fan obtained by optimization, represents the learning rate, Indicates the sensitivity of comfort to axial fan speed;

[0113] Based on the first optimization function, the second optimization function, the third optimization function and the parameter adaptive calibration function, the NSGA-III algorithm is used to solve the Pareto optimal solution set of the fluid simulation model to obtain the first fluid motion control parameter of the axial flow fan.

[0114] In this embodiment, the first optimization function aims to improve indoor comfort by minimizing the weighted sum of the predicted dissatisfaction percentage, the velocity gradient, and the sum of their squares. The predicted dissatisfaction percentage reflects the difference between indoor environmental parameters (such as temperature, humidity, and wind speed) and the user's desired comfort level. The velocity gradient and its sum of squares are used to assess the uniformity and stability of indoor airflow distribution. Excessive velocity gradients can lead to uneven airflow, affecting comfort.

[0115] The second optimization function focuses on the energy consumption of the axial fan and reduces the operating cost by minimizing the weighted sum of the axial fan speed and blade angle changes (taking into account their respective power coefficients converted to energy consumption).

[0116] The third optimization function aims to extend the service life of the axial fan. It balances equipment performance and service life by limiting the maximum allowable speed and considering material fatigue index and speed variance. The stability weighting factor is used to adjust the weight between speed and speed variance.

[0117] The parameter adaptive calibration function dynamically adjusts the learning rate based on the difference between the optimized axial fan speed and the actual operating speed, as well as the sensitivity of comfort to the fan speed, to achieve precise control of the fan speed. Through adaptive calibration, the system can dynamically adjust control parameters based on actual operating conditions and comfort requirements, improving control flexibility and accuracy. This helps further optimize energy consumption and equipment life while maintaining indoor comfort.

[0118] NSGA-III (Non-dominated Sorting Genetic Algorithm III) is a multi-objective optimization algorithm used to solve optimization problems involving multiple conflicting objectives. In this example, the NSGA-III algorithm is used to simultaneously optimize comfort, energy consumption, and equipment lifespan, thereby finding a set of Pareto-optimal solutions. These solutions achieve an optimal balance between the three objectives. Using the NSGA-III algorithm, a diverse set of control parameter combinations can be obtained that achieve an optimal trade-off between comfort, energy consumption, and equipment lifespan. Decision makers can select the most appropriate control strategy from these solutions based on actual needs to meet different application scenarios and requirements.

[0119] In some embodiments, in step S103, the step of constructing a personalized metabolic rate prediction model based on the human metabolic data of the indoor occupants and using the personalized metabolic rate prediction model to calculate the PMV-PPD thermal comfort index in real time specifically includes:

[0120] Collecting human metabolic data of indoor personnel, wherein the human metabolic data includes heart rate variability data, skin conductance data, and exercise intensity data;

[0121] constructing a multiple linear regression model based on the heart rate variability data, the skin conductance data, and the exercise intensity data, and calculating and obtaining a predicted metabolic rate through the multiple linear regression model;

[0122] The predicted metabolic rate is input into the PMV-PPD calculation model, and the PMV-PPD thermal comfort index is obtained as an output. The PMV-PPD calculation model satisfies

[0123]

[0124]

[0125] Among them, PMV represents the PMV index, PPD represents the percentage of unsatisfactory prediction, M represents the predicted metabolic rate, and W represents the work done by the human body. represents the water vapor partial pressure, Indicates the air temperature, represents the clothing area coefficient, represents the convective heat transfer coefficient, Indicates the surface temperature of clothing. represents the mean radiant temperature.

[0126] In this example, heart rate variability (HRV) refers to fluctuations in heart rate over a period of time, reflecting the activity of the autonomic nervous system. HRV data can be used to assess an individual's stress level, health status, and metabolic state. In building a predictive metabolic rate model, HRV data can serve as an important input variable due to its correlation with metabolic rate.

[0127] Skin conductance (SC) reflects the electrical conductivity of the skin and is commonly used to measure an individual's stress response. Changes in SC data are associated with emotional state, physiological arousal level, and metabolic activity. When predicting metabolic rate, SC data can serve as another useful input variable because it reflects an individual's physiological stress state and, therefore, indirectly reflects metabolic activity.

[0128] Exercise intensity refers to the level of energy expended during physical activity. Exercise intensity data can be obtained by measuring parameters such as heart rate, speed, and distance. Exercise intensity data is a key input variable in building predictive metabolic rate models because it directly affects an individual's metabolic rate.

[0129] Using a multivariate linear regression model, we can predict an individual's metabolic rate based on metabolic data (HRV, SC, and exercise intensity). This prediction method is fast and simple, and it can reflect an individual's actual metabolic state to a certain extent. Prediction accuracy can be improved by continuously optimizing model parameters.

[0130] The Predicted Mean Vote (PMV) index is a subjective indicator for evaluating thermal comfort. It takes into account multiple factors, including air temperature, humidity, wind speed, clothing thermal resistance, and the body's metabolic rate. PMV values ​​range from -3 to +3, with 0 representing the most comfortable state.

[0131] The PPD (Predicted Percentage of Dissatisfied) metric is short for predicted percentage of dissatisfaction. It indicates how many people are expected to be dissatisfied with their thermal comfort under given environmental conditions. The lower the PPD value, the more comfortable the environment.

[0132] By inputting the predicted metabolic rate into the PMV-PPD calculation model and combining it with other environmental parameters (such as water vapor partial pressure, air temperature, clothing area coefficient, convective heat transfer coefficient, clothing surface temperature, and mean radiant temperature), the PMV and PPD indices can be calculated. These indices are used to assess the thermal comfort level of indoor environments. By applying the PMV-PPD calculation model, the thermal comfort level of indoor environments can be quantitatively assessed, providing a scientific basis for improving indoor environments. Furthermore, combined with the predicted metabolic rate, it can more accurately reflect the thermal comfort needs of individuals in different activity states. This can help optimize building design, improve indoor environmental quality, and enhance people's life and work satisfaction.

[0133] In some embodiments, in the above step S103, the correction processing of the first fluid motion control parameter by using the PMV-PPD thermal comfort index to form the second fluid motion control parameter of the axial flow fan specifically includes:

[0134] The PMV-PPD thermal comfort index is input into the first control parameter correction function to obtain the axial fan speed adjustment value. The first control parameter correction function satisfies ,in, Indicates the speed adjustment of the axial fan. Indicates the percentage of dissatisfaction with the target, Indicates the current percentage of dissatisfaction, represents the proportional control coefficient, represents the integral control coefficient, represents the time integral of the PPD error;

[0135] The axial flow fan speed adjustment amount and the first fluid motion control parameter are added together to obtain a second fluid motion control parameter of the axial flow fan.

[0136] In this embodiment, the first control parameter correction function is designed based on the PID (Proportional-Integral-Derivative) control principle. It calculates the adjustment amount for the axial fan speed based on the difference between the current dissatisfaction percentage and the target dissatisfaction percentage. In this function, the target dissatisfaction percentage is preset and represents the desired indoor environmental comfort level; the current dissatisfaction percentage is obtained through the PMV-PPD calculation model and reflects the actual indoor comfort level. By inputting the PMV-PPD thermal comfort index into the first control parameter correction function, the axial fan speed can be precisely adjusted to improve indoor comfort. The PID control strategy comprehensively considers the current error (proportional term), the accumulation of past errors (integral term), and the error trend (although the differential term is not involved in this example, PID control often includes it), thereby achieving smooth and rapid adjustment of the controlled object.

[0137] The first fluid motion control parameter is derived using the previously presented fluid simulation model and optimization algorithm. It represents a set of first axial fan control parameters that balance multiple objectives (such as comfort, energy consumption, and equipment life). The axial fan speed adjustment is calculated based on the difference between the current indoor comfort level and the target level, reflecting the adjustment required to improve comfort. Adding these two parameters yields a new axial fan control parameter, the second fluid motion control parameter. This parameter considers both the balance between multiple objectives and the comfort requirements of the current environment. This method enables dynamic adjustment of axial fan control, enabling flexible adjustment of the control strategy based on the indoor environmental conditions at different times, thereby improving indoor comfort and energy efficiency.

[0138] In some embodiments, in the above step S104, the multimodal emotion recognition technology is used to analyze the user emotional state data of the indoor occupants, and the first fluid motion control parameter or the second fluid motion control parameter is corrected based on the user emotional state data to form the third fluid motion control parameter of the axial flow fan, specifically including:

[0139] Based on the voice modality data, visual modality data and physiological modality data of indoor occupants, the user's emotional state data is analyzed and output through a neural network model;

[0140] The user's emotional state data is input into the second control parameter correction function to obtain the axial fan speed correction value. The second control parameter correction function satisfies ,in, Indicates the axial fan speed correction value, Indicates the degree of arousal, Indicates the baseline arousal level, Indicates pleasure, Indicates baseline happiness, represents the adjustment coefficient of arousal, represents the adjustment coefficient of pleasure;

[0141] The axial flow fan speed correction value and the first fluid motion control parameter or the second fluid motion control parameter are added together to obtain a third fluid motion control parameter of the axial flow fan.

[0142] In this embodiment, the neural network model includes a feature encoding layer, an attention fusion layer and an output layer. The feature encoding layer satisfies , the attention fusion layer satisfies , , the output layer satisfies ,in, represents the encoded feature vector of the i-th mode, represents the eigenvector of the i-th mode, represents the long short-term memory network, represents the attention weight of the j-th modality, represents the encoded feature vector of the lth mode, represents the weight matrix of the lth mode, represents the weight matrix of the jth mode, Represents the feature vector after multimodal data fusion, Indicates the degree of arousal, Indicates pleasure, represents the weight matrix of the output layer, represents the bias vector of the output layer, Represents the tanh activation function.

[0143] The feature encoding layer encodes the voice, visual, and physiological modality data of the occupants into feature vectors. These feature vectors capture key information from each modality, providing the foundation for subsequent emotional state analysis. Through the feature encoding layer, data from different modalities is converted into unified feature vectors, facilitating subsequent processing and fusion. This enhances the system's flexibility and scalability, enabling it to handle a wide range of data inputs.

[0144] The attention fusion layer utilizes the attention mechanism to perform a weighted fusion of feature vectors from different modalities. Attention weights reflect the importance of different modalities in emotional state analysis, enabling the system to more accurately capture key information. Through the attention fusion layer, the system automatically adjusts the weights of different modalities, achieving effective fusion of multimodal data. This improves the accuracy and robustness of emotional state analysis, enabling the system to better adapt to different scenarios and user needs.

[0145] The output layer maps the fused feature vectors into the user's emotional state space, outputting two emotional state indicators: arousal and pleasure. These indicators reflect the user's emotional state and provide a basis for subsequent control parameter adjustments. Through the output layer, the system can convert multimodal data into specific emotional state indicators, providing a scientific basis for subsequent axial fan control. This helps achieve more personalized and intelligent indoor environmental control.

[0146] In this embodiment, the second control parameter correction function calculates the axial fan speed correction based on the difference between the user's emotional state data (arousal and pleasure) and the baseline value. The adjustment coefficients for arousal and pleasure determine the degree to which changes in emotional state affect the speed correction. Using this second control parameter correction function, the system can dynamically adjust the axial fan speed based on the user's emotional state, providing a more comfortable and personalized indoor environment. This helps improve user satisfaction and comfort while also achieving efficient energy utilization.

[0147] The axial fan speed correction is added to the first or second fluid motion control parameter to generate a third fluid motion control parameter. This parameter considers the balance between multiple objectives (such as comfort, energy consumption, and equipment lifespan) while also taking into account the user's emotional state. This approach enables the system to dynamically adjust axial fan control to adapt to the indoor environment and user's emotional state at different times. This helps improve indoor comfort and energy efficiency while meeting the user's personalized needs.

[0148] Reference Figure 2 An embodiment of the present invention provides a fluid motion control system 2 based on an axial flow fan, wherein the system 2 specifically includes:

[0149] The first control module 201 is configured to acquire indoor fluid state data and human motion data in real time using a sensor network, process the fluid state data and human motion data using a spatiotemporal alignment algorithm, and generate a joint representation model;

[0150] A second control module 202 is configured to construct a fluid simulation model based on the joint characterization model using LBM technology, and generate first fluid motion control parameters of the axial flow fan using the fluid simulation model and a preset multi-objective optimization function;

[0151] The third control module 203 is configured to construct a personalized metabolic rate prediction model based on the human metabolic data of the indoor occupant, calculate the PMV-PPD thermal comfort index in real time using the personalized metabolic rate prediction model, and modify the first fluid motion control parameter using the PMV-PPD thermal comfort index to form the second fluid motion control parameter of the axial flow fan;

[0152] The fourth control module 204 is used to use multimodal emotion recognition technology to analyze the user emotional state data of indoor people, and to correct the first fluid motion control parameter or the second fluid motion control parameter based on the user emotional state data to form the third fluid motion control parameter of the axial flow fan.

[0153] It is understandable that if Figure 1The contents of the embodiment of the fluid motion control method based on the axial flow fan shown in the figure are applicable to the embodiment of the fluid motion control system based on the axial flow fan. The functions specifically implemented by the embodiment of the fluid motion control system based on the axial flow fan are similar to those in the embodiment of the fluid motion control method based on the axial flow fan shown in the figure. Figure 1 The embodiment of the fluid motion control method based on the axial flow fan shown is the same as that of the embodiment shown in FIG. Figure 1 The beneficial effects achieved by the embodiment of the fluid motion control method based on the axial flow fan shown are also the same.

[0154] It should be noted that the information interaction, execution process and other contents between the above-mentioned systems are based on the same concept as the embodiment of the method of the present invention. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0155] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0156] Reference Figure 3 An embodiment of the present invention further provides a computer device 3, comprising: a memory 302 and a processor 301 and a computer program 303 stored in the memory 302. When the computer program 303 is executed on the processor 301, the fluid motion control method based on the axial flow fan as described in any one of the above methods is implemented.

[0157] The computer device 3 may be a desktop computer, a notebook computer, a PDA, a cloud server or other computing devices. The computer device 3 may include, but is not limited to, a processor 301 and a memory 302. Those skilled in the art will understand that Figure 3 This is merely an example of the computer device 3 and does not constitute a limitation on the computer device 3 . The computer device 3 may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the computer device 3 may also include input and output devices, network access devices, etc.

[0158] The processor 301 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. A general-purpose processor may be a microprocessor or any conventional processor.

[0159] In some embodiments, the memory 302 may be an internal storage unit of the computer device 3, such as a hard drive or memory of the computer device 3. In other embodiments, the memory 302 may also be an external storage device of the computer device 3, such as a plug-in hard drive, a Smart Media Card (SMC), a Secure Digital (SD) card, a flash memory card, etc. equipped on the computer device 3. Furthermore, the memory 302 may include both an internal storage unit of the computer device 3 and an external storage device. The memory 302 is used to store an operating system, application programs, a boot loader, data, and other programs, such as the program code of the computer program. The memory 302 may also be used to temporarily store data that has been output or is about to be output.

[0160] An embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for controlling fluid motion based on an axial flow fan as described in any one of the above methods is implemented.

[0161] In this embodiment, if the integrated unit is implemented as a software functional unit and sold or used as a standalone product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application can implement all or part of the process steps in the above-mentioned method embodiments by using a computer program to instruct the relevant hardware. The computer program can be stored in a computer-readable storage medium. When executed by a processor, the computer program can implement the steps of each of the above-mentioned method embodiments. The computer program includes computer program code, which can be in source code form, object code form, executable file, or some intermediate form. The computer-readable medium can include at least: any entity or device capable of carrying computer program code to a camera / terminal device, recording medium, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signals, telecommunication signals, and software distribution media. Examples include USB flash drives, removable hard drives, magnetic disks, or optical disks. In some jurisdictions, based on legislation and patent practice, computer-readable media cannot be electric carrier signals or telecommunication signals.

[0162] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0163] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0164] In the embodiments disclosed in the present application, it should be understood that the disclosed devices / terminal equipment and methods can be implemented in other ways. For example, the device / terminal equipment embodiments described above are merely schematic. For example, the division of the modules or units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, 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 through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0165] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

Claims

1. A fluid motion control method based on an axial flow fan, characterized in that: The method specifically includes: A sensor network is used to obtain indoor fluid state data and human motion data in real time. The fluid state data and human motion data are processed through a spatiotemporal alignment algorithm to generate a joint representation model. Based on the joint characterization model, a fluid simulation model is constructed using LBM technology, and the first fluid motion control parameters of the axial flow fan are generated through the fluid simulation model and a preset multi-objective optimization function; A personalized metabolic rate prediction model is constructed based on the human metabolic data of indoor occupants, a PMV-PPD thermal comfort index is calculated in real time using the personalized metabolic rate prediction model, and the first fluid motion control parameter is corrected using the PMV-PPD thermal comfort index to form a second fluid motion control parameter for the axial flow fan; Multimodal emotion recognition technology is used to analyze the user emotional state data of indoor occupants, and the first fluid motion control parameter or the second fluid motion control parameter is corrected based on the user emotional state data to form the third fluid motion control parameter of the axial flow fan.

2. The method according to claim 1, characterized in that The fluid state data and human motion data are processed by the spatiotemporal alignment algorithm to generate a joint representation model, specifically including: The SLAM technology is used to calibrate the spatial position of each sensor node in the room, and the extended Kalman filter is used to update the position of the self-moving sensor nodes in real time; An interpolation algorithm is used to unify the fluid state data and human motion data collected by the sensor network to the same time base; The room is divided into multiple grid cells, and the fluid state data and human motion data are mapped to the grid cells through Gaussian kernel density estimation to form a joint representation model.

3. The method according to claim 2, characterized in that The joint representation model satisfies in, represents the multidimensional feature vector of the grid unit (h, w, d) at time t, K represents the total number of sensors, represents the measurement value of the kth sensor, represents the spatial position of the kth sensor, represents the center coordinates of the grid cell (h,w,d), represents the spatial smoothing coefficient, represents the Euclidean distance between the sensor location and the center of the grid cell; in, represents the adaptive weight matrix used to adjust the contribution of sensor data to the joint representation model, represents the prediction joint representation model calculated by the weight matrix W, represents the true joint representation model obtained through experiments or high-precision measurements, represents the Frobenius norm of the weight matrix and is used to measure the difference between the predicted joint representation model and the true joint representation model, represents the regularization coefficient used to control the sparsity of the weight matrix, represents the L1 norm of the weight matrix.

4. The method according to claim 1, wherein The fluid simulation model is constructed based on the joint characterization model using LBM technology, specifically including: Map the grid cells of the joint representation model to the LBM computing network, and set the distribution function of each grid node of the LBM computing network ,in, It is used to represent the discrete velocity direction index of the LBM model, X=(x,y,z) is used to represent the grid node coordinates, and t represents the simulation time step; Set the axial fan outlet velocity boundary condition and dynamic obstacle boundary conditions ,in, represents the velocity field at the outlet of the axial fan, represents the velocity component of the axial fan outlet along the z-axis, Indicates the empirical coefficient used to convert the axial fan speed into the axial fan outlet speed. Indicates the speed of the axial flow fan, D indicates the diameter of the axial flow fan blade, represents the distribution function value at the boundary of a wall or obstacle, Represents the distribution function value in the direction opposite to i, represents the weight coefficient of the discrete velocity direction, represents the fluid density, The vector representing the direction of the i-th discrete velocity, represents the obstacle surface velocity provided by the joint characterization model, represents the speed of sound; The distribution function is initialized, and simulation is performed in combination with the axial flow fan air outlet velocity boundary condition and the dynamic obstacle boundary condition to form a fluid simulation model.

5. The method according to claim 4, characterized in that Initializing the distribution function specifically includes: When the distribution function At t=0, we get , , in, represents the distribution function value along the direction of the i-th discrete velocity at position X and time t=0, represents the initial fluid density, represents the initial velocity field, represents the equilibrium distribution function calculated from the initial fluid density and initial velocity field, represents the humidity data of the joint characterization model, Represents the wind speed data of the joint representation model.

6. The method according to claim 1, characterized in that Generating the first fluid motion control parameter of the axial flow fan by using the fluid simulation model and a preset multi-objective optimization function specifically includes: A first optimization function is set according to the comfort target, and the first optimization function satisfies ,in, represents the first optimization function value, t represents the simulation time step, T represents the total number of simulation time steps, represents the percentage of unsatisfactory predictions over time step t, represents the velocity gradient changing with time step t, represents the sum of squares of velocity gradients varying with time step t, represents the weight coefficient; A second optimization function is set according to the energy consumption target, and the second optimization function satisfies ,in, represents the second optimization function value, RPM(t) represents the axial fan speed that changes with time step t, represents the change in the axial fan blade angle as the time step t changes, Indicates the power factor used to convert the axial fan speed into energy consumption, Indicates the power coefficient used to convert the change in the axial fan blade angle into energy consumption; A third optimization function is set according to the equipment life target, and the third optimization function satisfies ,in, represents the third optimization function value, Indicates the maximum allowable speed of the axial fan. represents the material fatigue index, represents the speed variance of the axial fan, represents the stability weight coefficient used to balance the speed and speed variance of the axial fan; A parameter adaptive calibration function is set, and the parameter adaptive calibration function satisfies ,in, Indicates the adjusted axial fan speed, represents the optimal speed of the axial fan obtained by optimization, represents the learning rate, Indicates the sensitivity of comfort to axial fan speed; Based on the first optimization function, the second optimization function, the third optimization function and the parameter adaptive calibration function, the NSGA-III algorithm is used to solve the Pareto optimal solution set of the fluid simulation model to obtain the first fluid motion control parameter of the axial flow fan.

7. The method according to claim 1, characterized in that The personalized metabolic rate prediction model is constructed based on the human metabolic data of indoor occupants, and the PMV-PPD thermal comfort index is calculated in real time using the personalized metabolic rate prediction model, specifically including: Collecting human metabolic data of indoor personnel, wherein the human metabolic data includes heart rate variability data, skin conductance data, and exercise intensity data; constructing a multiple linear regression model based on the heart rate variability data, the skin conductance data, and the exercise intensity data, and calculating and obtaining a predicted metabolic rate through the multiple linear regression model; The predicted metabolic rate is input into the PMV-PPD calculation model, and the PMV-PPD thermal comfort index is obtained as an output. The PMV-PPD calculation model satisfies Among them, PMV represents the PMV index, PPD represents the percentage of unsatisfactory prediction, M represents the predicted metabolic rate, and W represents the work done by the human body. represents the water vapor partial pressure, Indicates the air temperature, represents the clothing area coefficient, represents the convective heat transfer coefficient, Indicates the surface temperature of clothing. represents the mean radiant temperature.

8. The method according to claim 1, characterized in that The correcting process of the first fluid motion control parameter by using the PMV-PPD thermal comfort index to form the second fluid motion control parameter of the axial flow fan specifically includes: The PMV-PPD thermal comfort index is input into the first control parameter correction function to obtain the axial fan speed adjustment value. The first control parameter correction function satisfies ,in, Indicates the speed adjustment of the axial fan. Indicates the percentage of dissatisfaction with the target, Indicates the current percentage of dissatisfaction, represents the proportional control coefficient, represents the integral control coefficient, represents the time integral of the PPD error; The axial flow fan speed adjustment amount and the first fluid motion control parameter are added together to obtain a second fluid motion control parameter of the axial flow fan.

9. The method according to claim 1, characterized in that The multimodal emotion recognition technology is used to analyze the user emotional state data of the indoor occupants, and the first fluid motion control parameter or the second fluid motion control parameter is corrected based on the user emotional state data to form the third fluid motion control parameter of the axial flow fan, specifically including: Based on the voice modality data, visual modality data and physiological modality data of indoor occupants, the user's emotional state data is analyzed and output through a neural network model; The user's emotional state data is input into the second control parameter correction function to obtain the axial fan speed correction value. The second control parameter correction function satisfies ,in, Indicates the axial fan speed correction value, Indicates the degree of arousal, Indicates the baseline arousal level, Indicates pleasure, Indicates baseline happiness, represents the adjustment coefficient of arousal, represents the adjustment coefficient of pleasure; The axial flow fan speed correction value and the first fluid motion control parameter or the second fluid motion control parameter are added together to obtain a third fluid motion control parameter of the axial flow fan.

10. A fluid motion control system based on an axial flow fan, characterized in that: The system specifically includes: The first control module is used to use a sensor network to obtain indoor fluid state data and human motion data in real time, process the fluid state data and human motion data through a spatiotemporal alignment algorithm, and generate a joint representation model; A second control module is configured to construct a fluid simulation model based on the joint characterization model using LBM technology, and generate first fluid motion control parameters of the axial flow fan using the fluid simulation model and a preset multi-objective optimization function; a third control module, configured to construct a personalized metabolic rate prediction model based on the human metabolic data of the indoor occupants, calculate the PMV-PPD thermal comfort index in real time using the personalized metabolic rate prediction model, and correct the first fluid motion control parameter using the PMV-PPD thermal comfort index to form a second fluid motion control parameter for the axial flow fan; The fourth control module is used to use multimodal emotion recognition technology to analyze the user emotional state data of indoor people, and to correct the first fluid motion control parameter or the second fluid motion control parameter based on the user emotional state data to form the third fluid motion control parameter of the axial flow fan.

Images