Intelligent adjustment method for air supply flow organization of high and large space air conditioner

Abstract

The application relates to the field of air conditioning and intelligent control, and discloses an intelligent adjustment method for air supply airflow organization in a high and large space. The method comprises the following steps: obtaining a three-dimensional thermal environment field, a personnel space distribution density and air conditioner operation parameters; constructing a space thermal comfort evaluation model which integrates a vertical temperature gradient, a local heat load and a personnel heat-sensitive area; generating air supply angle and air supply amount adjustment instructions according to the space thermal comfort evaluation model, and driving a variable frequency fan to perform dynamic regulation and control. The intelligent adjustment mechanism constructed by the application has strong robustness and self-adaptability, and is suitable for various high and large space building types such as stadiums, exhibition centers and airport terminals, and solves the technical problem of long-existing airflow organization and personnel demand mismatch.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of heating, ventilation, air conditioning and intelligent control, and particularly relates to an intelligent adjustment method for air supply airflow organization in a high and large space. BACKGROUND

[0002] High and large space buildings such as stadiums, exhibition centers and airport terminals have a significant vertical height and a large internal volume, and there is a significant thermal stratification phenomenon. During heating or refrigeration, hot air naturally rises and gathers in the top area due to its lower density, while cold air sinks to the personnel activity area, resulting in a serious uneven temperature distribution in the vertical direction.

[0003] The traditional air conditioning system airflow organization strategy is based on static design working conditions, and lacks the ability to perceive and respond to dynamic heat load changes in the space. Especially when there is a sudden increase in instantaneous heat load in the personnel-intensive area, the fixed air supply mode cannot adjust the airflow direction and intensity in time, which easily causes the cold air to fail to effectively reach the target area, or the hot air to be retained in the upper space, not only reducing human thermal comfort, but also possibly causing local overcooling or overheating, increasing energy consumption and operating costs.

[0004] The core of airflow organization optimization of the air conditioning system in a high and large space is to achieve precise coverage and dynamic adaptation of the air supply airflow to the personnel activity area. In an ideal state, the air supply angle should be automatically adjusted according to the real-time thermal environment changes in the space to overcome the airflow deviation caused by thermal buoyancy effect, and to ensure that the cold air or hot air efficiently reaches the target area. However, the existing technology generally relies on preset programs or manual intervention for mechanical louver adjustment, which has a slow response and low adjustment accuracy, and is difficult to cope with complex and variable personnel distribution and external climate disturbances.

[0005] In the existing technology, some systems attempt to introduce temperature sensing probes or infrared sensors for local temperature monitoring, but their sparse distribution and single sensing dimension cannot construct a complete three-dimensional temperature field model. At the same time, the control logic used is mostly threshold trigger type on-off control, which lacks the ability to predict the evolution trend of temperature gradient, resulting in frequent adjustment actions or slow response. Traditional adjustment mechanisms mostly rely on step motor driven louvers, which have complex mechanical structures and slow response speed, and are difficult to achieve millisecond level angle optimization in the rapidly changing thermal environment of a high and large space.

[0006] The above defects collectively result in the existing air supply system being unable to effectively suppress the problems of insufficient cold air sinking or hot air retention when facing local heat load mutation, causing a high rate of decline in personnel comfort, and there is an urgent need for an airflow organization intelligent adjustment method that integrates high-density perception, lightweight intelligent prediction and fast execution mechanism. SUMMARY

[0007] The application provides a high and large space air conditioning air supply flow organization intelligent adjustment method, and aims to solve the problem of static setting defects of air supply angle caused by thermal stratification effect in high and large space.

[0008] The application realizes real-time, accurate and self-adaptive regulation and control of air supply flow organization by constructing an intelligent air flow adjustment system that integrates multi-source perception, thermal environment modeling, personnel dynamic identification and closed-loop control of execution mechanism.

[0009] The high and large space air conditioning air supply flow organization intelligent adjustment method provided by the application comprises the following steps: obtaining three-dimensional thermal environment field data, personnel spatial distribution density data and current operation parameters of an air conditioning system in a high and large space;

[0010] Based on the three-dimensional thermal environment field data and the personnel spatial distribution density data, a spatial thermal comfort evaluation model containing vertical temperature gradient, local thermal load intensity and personnel heat sensitive area is constructed;

[0011] According to the output result of the spatial thermal comfort evaluation model, air supply angle adjustment instructions and air supply amount distribution instructions for each air supply port are generated;

[0012] The air supply angle adjustment instructions and the air supply amount distribution instructions are issued to the corresponding variable frequency fan controller to complete dynamic adjustment of the air supply flow organization.

[0013] Further, obtaining the three-dimensional thermal environment field data in the high and large space specifically comprises:

[0014] Air temperature values of each measuring point are collected through a distributed temperature sensor array arranged at different height layers inside the high and large space;

[0015] The distributed temperature sensor array is divided into not less than five height layers along the vertical direction, and is uniformly arranged in a grid shape in each layer;

[0016] The surface temperature field of the main personnel activity area in the high and large space is scanned non-contactly by an infrared thermal imager to obtain the radiation temperature distribution of the wall, seat, ground and human body surface;

[0017] Combining the air temperature data and the radiation temperature data, a continuous three-dimensional temperature field model is reconstructed by using the inverse distance weighted interpolation algorithm.

[0018] Further, obtaining the personnel spatial distribution density data specifically comprises:

[0019] Real-time video image acquisition of the whole field is performed by a wide-angle visible light camera array installed at the top of the high and large space;

[0020] The video image is foreground segmented and human target detected, and a target detection model based on deep learning is used to identify the human contour in the image;

[0021] According to the pixel position of the human contour in the image and the known camera internal and external parameters, a two-dimensional image coordinate is mapped to a three-dimensional space coordinate through perspective transformation;

[0022] The number of effective human targets in a unit volume space is counted, and a personnel space density distribution voxel map in cubic meters is generated;

[0023] The spatial resolution of the voxel map is 2*2*1 meters.

[0024] Further, obtaining the current operating parameters of the air conditioning system specifically includes: reading the current air volume, static pressure, supply air temperature and fan speed of each air supply branch from the building automation system interface; reading the opening state and feedback signal of each electric regulating valve; reading the linkage state of the fresh air valve and the return air valve; the sampling period of the operating parameters is 5 seconds.

[0025] Further, constructing a space thermal comfort evaluation model including vertical temperature gradient, local heat load intensity and personnel heat sensitive area specifically includes:

[0026] The temperature difference between any two adjacent height layers is calculated to form a vertical temperature gradient vector; the personnel dense area is defined as a voxel set with a personnel space density greater than 0.5 person per cubic meter; in the personnel dense area, the absolute value of the deviation of the average temperature of the area from the set comfortable temperature is calculated;

[0027] The absolute value of the temperature deviation is weighted and summed with the vertical temperature gradient within two meters above the area, and the weight coefficients are 0.7 and 0.3 respectively, to obtain a local thermal discomfort index;

[0028] All personnel dense areas are traversed, and areas with a local thermal discomfort index greater than a preset threshold of 1.5 degrees Celsius per meter are selected and marked as heat sensitive regulation areas.

[0029] Further, generating air supply angle adjustment instructions and air supply volume distribution instructions for each air supply port specifically includes:

[0030] A spatial mapping relationship table of air supply ports and heat sensitive regulation areas is established, and the mapping relationship table is pre-calibrated according to the physical position of the air supply port, the jet coverage range and the air flow attenuation characteristics;

[0031] For each heat sensitive regulation area, the associated air supply port set is queried; according to the vertical position of the heat sensitive regulation area, the initial adjustment direction of the air supply angle is determined:

[0032] If the region is located in the lower 1 / 3 height of the space, the air supply angle is deflected downward; if located in the middle, it remains horizontal or slightly adjusted;

[0033] If located in the upper part, it is deflected upward to suppress the accumulation of hot air; the specific deflection amount of the air supply angle is obtained by linear mapping of the local thermal discomfort index, and the mapping relationship is , is the deflection angle, is the local thermal discomfort index, and the maximum deflection angle is less than ±30 degrees;

[0034] At the same time, according to the personnel density and area in the heat-sensitive regulation area, the air supply amount of the associated air supply outlet in the area is increased in proportion, and the increment is the reference air volume multiplied by the ratio of the personnel density to 0.5 person per cubic meter, and the maximum increment is less than 50% of the reference air volume.

[0035] Further, the variable frequency fan controller receives the air supply amount distribution instruction, and changes the fan rotating speed by adjusting the power supply frequency of the fan motor;

[0036] The fan is a centrifugal fan, and the rated air volume is 50,000 cubic meters per hour, and the rated total pressure is 800 Pa;

[0037] The frequency conversion range is 25 Hz to 50 Hz, corresponding to the air volume adjustment range of 50% to 100%; the air volume adjustment response time is less than 30 seconds.

[0038] Further, after completing the dynamic adjustment of the air supply air flow organization once, the system enters a continuous monitoring state, and the whole process of the above data acquisition, model construction, instruction generation and execution control is repeatedly executed every 15 seconds, forming a closed-loop adaptive adjustment mechanism;

[0039] When the local thermal discomfort index of the heat-sensitive regulation area is less than the preset threshold value after three consecutive adjustments, the angle adjustment is suspended, and only the air volume adjustment is maintained;

[0040] When it is detected that the personnel distribution has changed significantly, that is, the change rate of the personnel density distribution voxel graph is greater than a specified value of 20%, a new round of adjustment process is triggered immediately.

[0041] As a preferred embodiment of the present application, the working waveband of the infrared thermal imager is 8 to 14 microns, the spatial resolution is 640×480 pixels, and the temperature measurement accuracy is ±0.5 degrees Celsius; the field of view angle of the wide-angle visible light camera is not less than 120 degrees, the resolution is 1920×1080 pixels, and the frame rate is 25 frames per second; the distributed temperature sensor uses a platinum resistance temperature sensor, the accuracy level is A level, and the measurement range is zero to 50 degrees Celsius.

[0042] As a preferred embodiment of the present application, the deep learning-based target detection model adopts an improved YOLOv5 architecture, the backbone network of which is replaced with a lightweight MobileNetV3, the training data set contains no less than 100,000 high and large space scene images labeled with human body targets, and the average precision mean of the model on the test set is greater than 92%.

[0043] As a preferred embodiment of the present application, the power exponent parameter in the inverse distance weighted interpolation algorithm is set to 2, and the search radius is set to 15 meters; the preset threshold 1.5 degrees Celsius per meter in the space thermal comfort evaluation model is a critical value obtained by statistical analysis of thermal sensation voting experiments of no less than 500 subjects under different vertical temperature gradients.

[0044] As a preferred embodiment of the present application, the jet coverage range of the air supply outlet is pre-calibrated by CFD numerical simulation, the simulation working conditions include summer cooling and winter heating modes, the air supply speed range is 2 to 6 meters per second, the total number of simulation grids is no less than 5 million, and the turbulent flow model adopts the Realizable k-epsilon model.

[0045] Compared with the prior art, the present application has the following advantages:

[0046] 1. The present application realizes dynamic, accurate and closed-loop regulation and control of air supply flow organization in high and large space by fusing three-dimensional thermal environment perception, high-precision personnel distribution identification and thermal comfort quantitative evaluation.

[0047] 2. The present application can respond to local thermal load mutation in personnel dense area in real time, actively adjust air supply angle and air volume, suppress thermal stratification effect, and improve cold air delivery effect in lower personnel activity area or upper hot air retention problem.

[0048] 3. The intelligent adjustment mechanism constructed by the present application has strong robustness and adaptability, and is suitable for various high and large space building types such as stadiums, exhibition centers and airport terminals, solving the long-existing technical problem of mismatch between air flow organization and personnel demand. BRIEF DESCRIPTION OF DRAWINGS

[0049] Figure 1 is the overall technical scheme architecture schematic diagram of the intelligent adjustment method of air supply flow organization in high and large space air conditioning proposed by the present application;

[0050] Figure 2 is the core principle framework schematic diagram of the space thermal comfort evaluation model in the present application;

[0051] Figure 3 is the logic flow framework diagram of three-dimensional thermal environment field and personnel space distribution density data fusion perception in the present application;

[0052] Figure 4 is the logical flow framework diagram of the air supply angle and air supply volume dynamic generation and instruction issuing in the present application;

[0053] Figure 5 is a multi-level interaction relationship and data flow diagram of the system adaptive adjustment cycle and triggering mechanism in the present application. DETAILED DESCRIPTION

[0054] Embodiment 1: Please refer to Figures 1 to 5 The present application provides an intelligent adjustment method for air supply airflow organization of high and large space air conditioning, aiming to solve the problem of static setting defect of air supply angle caused by thermal stratification effect.

[0055] In high and large space buildings such as stadiums, theaters, audience seats or large exhibition halls, the traditional air conditioning system generally uses mechanical air supply outlets with preset angles, which cannot dynamically adjust the air supply direction and airflow intensity according to the personnel distribution, local heat load change and indoor thermal environment, resulting in insufficient cold air sinking, hot air stagnating in the upper space, and large vertical temperature gradient, etc., which further leads to the decline of human thermal comfort.

[0056] The present embodiment realizes real-time, accurate and adaptive regulation and control of air supply airflow organization by constructing an intelligent airflow adjustment system integrating multi-source perception, thermal environment modeling, dynamic personnel identification and closed-loop control of execution mechanism.

[0057] The intelligent adjustment method for air supply airflow organization of high and large space air conditioning comprises the following steps:

[0058] Obtain three-dimensional thermal environment field data, personnel spatial distribution density data and current operation parameters of the air conditioning system in the high and large space;

[0059] Based on the three-dimensional thermal environment field data and personnel spatial distribution density data, a spatial thermal comfort evaluation model is constructed, which includes vertical temperature gradient, local heat load intensity and personnel heat sensitive area;

[0060] According to the output result of the spatial thermal comfort evaluation model, air supply angle adjustment instructions and air supply volume distribution instructions are generated for each air supply outlet;

[0061] The air supply angle adjustment instructions and air supply volume distribution instructions are issued to the corresponding variable frequency fan controller to complete the dynamic adjustment of air supply airflow organization.

[0062] The three-dimensional thermal environment field data in the high and large space are obtained, which specifically includes:

[0063] Through the deployment of distributed temperature sensor array at different height layers inside the high and large space, the air temperature values of each measuring point are collected;

[0064] The distributed temperature sensor array is divided into not less than five height layers in the vertical direction, and sensors are uniformly arranged in a grid in each layer.

[0065] The infrared thermal imager is installed on the top structure of the high and large space, and its field of view covers all the personnel-intensive areas. The working waveband is 8-14 microns, the spatial resolution is 640x480 pixels, and the temperature measurement accuracy is ±0.5 degrees Celsius.

[0066] The infrared thermal imager scans the target area at a frequency of one frame per second to obtain the surface radiation temperature distribution.

[0067] The distributed temperature sensor array is divided into not less than five height layers in the vertical direction, and sensors are uniformly arranged in a grid in each layer.

[0068] In each height layer, the sensors are arranged in a 10x10 meter square grid to ensure that the horizontal distance from any position to the nearest sensor is less than 7.07 meters.

[0069] All sensors use platinum resistance temperature sensors with an accuracy level of A, a measurement range of 0-50 degrees Celsius, a sampling period of 5 seconds, and real-time data upload to the central controller through wired or wireless means.

[0070] The infrared thermal imager is installed on the top structure of the high and large space, and its field of view covers all the personnel-intensive areas. The working waveband is 8-14 microns, the spatial resolution is 640x480 pixels, and the temperature measurement accuracy is ±0.5 degrees Celsius.

[0071] The infrared thermal imager scans the target area at a frequency of one frame per second to obtain the surface radiation temperature distribution.

[0072] The infrared thermal imager scans the target area at a frequency of one frame per second to obtain the surface radiation temperature distribution.

[0073] After obtaining the air temperature point data and the surface radiation temperature field, the central controller performs the inverse distance weighted interpolation algorithm to reconstruct the continuous three-dimensional temperature field.

[0074] The algorithm takes each known measurement point as a reference and performs a weighted average calculation on the temperature value of any unknown point in space, with the weight being inversely proportional to the square of the Euclidean distance from the point to the unknown point.

[0075] The power exponent parameter in the algorithm is set to 2, and the search radius is set to 15 meters, i.e., only considering the influence of measurement points within a 15-meter range.

[0076] Through this interpolation process, a three-dimensional temperature voxel model with a resolution of 1x1x1 meter is generated as the basis input for subsequent thermal comfort evaluation.

[0077] The obtaining personnel spatial distribution density data specifically includes:

[0078] A wide-angle visible light camera array installed on the top of the high space is used to collect real-time video images of the whole field.

[0079] The video images are subjected to foreground segmentation and human body target detection, and a target detection model based on deep learning is used to identify the human body contour in the image.

[0080] According to the pixel position of the human body contour in the image and the known camera internal and external parameters, the two-dimensional image coordinates are mapped to three-dimensional space coordinates through perspective transformation.

[0081] The number of effective human body targets in a unit volume of space is counted to generate a personnel spatial density distribution voxel map in cubic meters.

[0082] The spatial resolution of the voxel map is 2x2x1 meters.

[0083] The wide-angle visible light camera array is composed of multiple cameras, each with a field of view angle of not less than 120 degrees, a resolution of 1920x1080 pixels, and a frame rate of 25 frames per second.

[0084] The camera installation position is calibrated geometrically to ensure that there is no visual blind area in the whole field. The video stream is transmitted in real time to the central controller, and the foreground segmentation is performed by the embedded image processing unit. The moving target is separated by using the mixed Gaussian background modeling method.

[0085] Subsequently, a target detection model based on deep learning is called to identify human bodies in the foreground area.

[0086] The model uses an improved YOLOv5 architecture, with a lightweight MobileNetV3 backbone network. The training data set contains not less than 100,000 high space scene images annotated with human body targets. The average precision of the model on the test set is greater than 92%.

[0087] After target detection, each identified human body target outputs its two-dimensional bounding box center coordinates in the image.

[0088] Using the pre-calibrated camera internal and external parameter matrices (including rotation matrix and translation vector), the two-dimensional coordinates are mapped to three-dimensional world coordinates through inverse perspective projection transformation.

[0089] During the mapping process, the height of the human body is assumed to be 1.7 meters, and the projection position of the feet on the ground is estimated accordingly, and the height in the vertical direction is calculated in combination with the camera viewing angle.

[0090] The three-dimensional coordinates of all valid human body targets are recorded and spatially divided into voxel units of 2*2*1 meters.

[0091] The number of people in each voxel is divided by its volume to obtain the personnel spatial density value of the voxel, with the unit being people per cubic meter.

[0092] Finally, a personnel density distribution voxel map covering the entire space is formed, with an update period of 5 seconds.

[0093] The acquisition of the current operating parameters of the air conditioning system specifically includes:

[0094] Reading the current air volume, static pressure, supply air temperature, and fan speed of each air supply branch from the building automation system interface;

[0095] Reading the opening state and feedback signal of each electric regulating valve;

[0096] Reading the linkage state of the fresh air valve and the return air valve; the sampling period of the operating parameters is 5 seconds.

[0097] All parameters are obtained in real time from the building automation system through standard communication protocols such as BACnet or ModbusTCP, and stored in the operating state buffer area of the central controller for subsequent instruction generation module calls.

[0098] After obtaining the above three types of data, the construction of the space thermal comfort evaluation model is performed.

[0099] The model specifically includes:

[0100] Calculating the temperature difference between any two adjacent height layers to form a vertical temperature gradient vector;

[0101] Defining a personnel-intensive area as a set of voxels with a personnel spatial density greater than 0.5 people per cubic meter;

[0102] In the personnel-intensive area, calculate the absolute value of the deviation of the average temperature of the area from the set comfortable temperature;

[0103] Weighted sum of the temperature deviation absolute value and the vertical temperature gradient within the two-meter range above the area, with weight coefficients of 0.7 and 0.3 respectively, to obtain the local thermal discomfort index;

[0104] Traverse all personnel-intensive areas and select areas with a local thermal discomfort index greater than a preset threshold of 1.5 degrees Celsius per meter, marked as heat-sensitive control areas.

[0105] The calculation of vertical temperature gradient vector is based on the three-dimensional temperature field model. For any horizontal position, its temperature values at five height layers are extracted, and the temperature difference between adjacent layers is calculated in turn and divided by the height difference between layers to obtain four local gradient values. The gradient vector is used to represent the thermal stratification strength of the position.

[0106] The set comfort temperature is 26 degrees Celsius by default or preset by the user.

[0107] For each voxel with a person density greater than 0.5 person per cubic meter, a local area is formed by aggregating its adjacent voxels centered on it, and the arithmetic mean of the temperatures of all voxels in the area is calculated, and then the absolute deviation from the set comfort temperature is calculated.

[0108] At the same time, the maximum vertical temperature gradient value in the two-meter height range directly above the area is extracted. The local thermal discomfort index The following formula is used for calculation: ;

[0109] is the average temperature of the local area, is the set comfort temperature, is the maximum vertical temperature gradient in the two-meter range above, in degrees Celsius per meter.

[0110] The preset threshold of 1.5 degrees Celsius per meter is a critical value obtained by statistical analysis of the thermal sensation voting experiment of no less than 500 subjects under different vertical temperature gradients.

[0111] All areas with LTI greater than the threshold are marked as heat-sensitive regulation zones, and their spatial positions, LTI values, person densities, and vertical height ranges are recorded.

[0112] After determining the heat-sensitive regulation zones, the generation of air supply angle adjustment instructions and air supply amount distribution instructions is performed.

[0113] The process specifically includes:

[0114] A spatial mapping relationship table between air supply outlets and heat-sensitive regulation zones is established, and the mapping relationship table is pre-calibrated according to the physical positions of the air supply outlets, the jet coverage range, and the air flow attenuation characteristics;

[0115] For each heat-sensitive regulation zone, the associated set of air supply outlets is queried; and according to the vertical position of the heat-sensitive regulation zone, the initial adjustment direction of the air supply angle is determined:

[0116] If the area is located in the lower 1 / 3 height of the space, the air supply angle is deflected downward;

[0117] If it is located in the middle, it is kept horizontal or fine-tuned;

[0118] If located in the upper part, deflect upward to suppress the accumulation of hot air;

[0119] The specific deflection amount of the air supply angle is linearly mapped from the local thermal discomfort index, with the mapping relationship being , is the deflection angle, is the local thermal discomfort index, and the maximum deflection angle is less than ±30 degrees;

[0120] At the same time, according to the personnel density and area in the heat-sensitive regulation area, the air supply amount of the associated air supply outlet in this area is proportionally increased, and the increment is the reference air volume multiplied by the ratio of the personnel density to 0.5 person per cubic meter, and the maximum increment is less than 50% of the reference air volume.

[0121] The spatial mapping relationship table of the air supply outlet and the heat-sensitive regulation area is pre-calibrated through CFD numerical simulation in the system initialization stage.

[0122] The simulation conditions include two modes of summer refrigeration and winter heating, the air supply speed range is 2 to 6 meters per second, the total number of simulation grids is not less than 5 million, and the turbulence model adopts the Realizable k-epsilon model.

[0123] The jet trajectory, speed decay curve and coverage volume of each air supply outlet are accurately calculated and stored as a three-dimensional Boolean mask.

[0124] When the position of the heat-sensitive regulation area is determined, the system queries which air supply outlets have spatial intersection with the jet coverage range, thereby determining the associated air supply outlet set.

[0125] The adjustment direction of the air supply angle is determined according to the vertical position of the heat-sensitive regulation area. Assuming that the total net height of the large space is H, the lower area is defined as 0 to H / 3, the middle area is H / 3 to 2H / 3, and the upper area is 2H / 3 to H. If the center of gravity of the heat-sensitive regulation area is located in the lower part, the corresponding air supply outlet louver is deflected downward to enhance the cold air sinking effect; if it is located in the upper part, it is deflected upward to avoid hot air blowing directly to the personnel, while promoting the circulation of hot air in the upper part.

[0126] The deflection angle is calculated according to the following formula: ;

[0127] The unit is degrees Celsius per meter, The unit is degrees, and The sign is determined by the area position (negative for the lower part and positive for the upper part).

[0128] This angle instruction is packaged as the number of pulses.

[0129] The air supply amount distribution instruction is based on the personnel density and the reference air volume Calculation. Increment But Final target air volume The air volume instruction is converted into a fan frequency instruction and sent to the variable frequency fan controller.

[0130] The variable frequency fan controller receives the air supply volume distribution instruction and changes the fan speed by adjusting the power supply frequency of the fan motor;

[0131] The fan is a centrifugal fan with a rated air volume of 50,000 cubic meters per hour and a rated total pressure of 800 Pa;

[0132] The frequency conversion range is 25 Hz to 50 Hz, corresponding to an air volume adjustment range of 50% to 100%; the air volume adjustment response time is less than 30 seconds.

[0133] The fan air volume and frequency are approximately linearly related, the controller has a built-in lookup table method to map the target air volume to an accurate frequency value, and a PID algorithm is used to suppress speed fluctuations.

[0134] After completing a dynamic adjustment of the air supply flow organization, the system enters a continuous monitoring state, repeating the above data acquisition, model construction, instruction generation and execution control process every 15 seconds to form a closed-loop adaptive adjustment mechanism;

[0135] When the local thermal discomfort index of the heat-sensitive control area is less than the preset threshold value after three consecutive adjustments, the angle adjustment is suspended and only air volume fine-tuning is maintained;

[0136] When a significant change in personnel distribution is detected, i.e., the change rate of the personnel density distribution voxel map is greater than the specified value of 20%, a new round of adjustment process is triggered immediately.

[0137] The change rate calculation method is:

[0138] Compare the current voxel map with the previous cycle voxel map, calculate the proportion of the sum of the density change absolute values to the total volume, and if the proportion is greater than 20%, it is considered to be a significant change.

[0139] The intelligent adjustment mechanism constructed in this embodiment has strong robustness and adaptability, and is suitable for various types of large space buildings.

[0140] Through the above method, the vertical temperature gradient can be controlled within one to two degrees Celsius per meter, the thermal comfort compliance rate of the personnel-intensive area is improved to more than 95%, the air conditioning system energy consumption is reduced by 15% to 20%, and the discomfort caused by excessive air supply is avoided.

[0141] The method realizes the whole-link closed-loop control from environment perception, demand identification, strategy generation to execution feedback, and solves the long-existing technical problem of airflow organization and personnel demand mismatch.

[0142] At the system level, the hardware platform relied on by the embodiment includes a distributed temperature sensor array, an infrared thermal imager, a wide-angle visible light camera array, a central controller, and a variable frequency fan controller.

[0143] The central controller is an industrial-grade embedded computer equipped with a multi-core processor and a dedicated AI acceleration module, running a real-time operating system to ensure that each task is executed strictly on time.

[0144] All sensors and actuators are connected through industrial Ethernet, and the communication protocol uses time-sensitive network technology to ensure the determinism and timeliness of command transmission.

[0145] The variable frequency fan controller has a local fault diagnosis function, which can maintain the last valid command state when communication is interrupted and report an exception code.

[0146] The entire system supports remote monitoring and parameter configuration, and can be integrated with a higher-level energy management system through a standard API to achieve higher-level optimization scheduling.

[0147] Embodiment 2: The data acquisition, model construction, instruction generation, execution feedback, and closed-loop optimization whole-link architecture of Embodiment 1 is followed, and the core modules include: multi-source data perception, spatial thermal comfort evaluation model, air supply parameter adjustment instruction generation, execution mechanism closed-loop control, and adaptive adjustment mechanism. Similarly, to solve the problem of airflow organization and personnel demand mismatch caused by thermal stratification in high and large spaces, to achieve precise air supply coverage and dynamic adaptation, to suppress the phenomenon of hot air accumulation or insufficient cold air sinking, to improve personnel thermal comfort, and to reduce air conditioning system energy consumption.

[0148] This embodiment adapts to a rotational flow induction type side air supply stratification air conditioning system, and realizes dynamic airflow optimization in high and large spaces by integrating rotational flow air supply characteristics and intelligent control logic. The specific steps are as follows:

[0149] Step 1, data acquisition: adapt to rotational flow air supply system to perceive demand:

[0150] Three-dimensional thermal environment field data acquisition:

[0151] Through the deployment of a distributed temperature sensor array at different height layers inside the high and large space, the air temperature values of each measurement point are collected; the sensor array is divided into not less than five height layers along the vertical direction, and each layer is uniformly arranged in a grid shape (grid spacing not greater than 10 meters), covering the jet coverage range of the rotational flow air supply port.

[0152] The infrared thermal imager is used to scan the main personnel activity area non-contact, and the wall, floor, seat and human body surface radiation temperature distribution is obtained. The temperature field deviation under the influence of the rotational flow air current is corrected.

[0153] Combined with air temperature and radiation temperature data, the inverse distance weighted interpolation algorithm (power index = 2, search radius = 15 meters) is used to reconstruct the continuous three-dimensional temperature field model, with a resolution of 1×1×1 meters, and the temperature change in the rotational flow stratification area is accurately captured.

[0154] Personnel space distribution density data acquisition:

[0155] Real-time video images are collected by high and large space top wide-angle visible light camera array (field of view angle ≥120°, resolution 1920×1080 pixels) covering the main service area of the rotational flow system.

[0156] The human body contour is identified by foreground segmentation and improved YOLOv5 target detection model (MobileNetV3 as the backbone network), which is mapped to three-dimensional space coordinates through perspective transformation, and a spatial resolution of 2×2×1 meter personnel density distribution voxel map is generated (update period 5 seconds).

[0157] The personnel dense area is defined as the voxel set with density > 0.5 person / m3, and the rotational flow induction coverage range of the rotational flow air supply outlet is matched.

[0158] Air conditioning system operating parameter acquisition:

[0159] The current air volume, static pressure, supply air temperature and fan speed of each air supply branch are read from the building automation system interface, with a sampling period of 5 seconds.

[0160] The linkage state of the rotational flow air supply outlet, the feedback signal of the electric regulating valve opening degree, and the cooperative operation parameters of the fresh air valve and the return air valve are read to ensure the adaptation to the intelligent adjustment instruction.

[0161] Step 2, construction of space thermal comfort evaluation model:

[0162] The temperature difference of adjacent height layers is calculated to form a vertical temperature gradient vector, and the gradient change in the rotational flow stratification air supply area is analyzed (target control ≤1.2℃ / m).

[0163] In the personnel dense area, the absolute value of the deviation of the average temperature from the set comfortable temperature (default 26℃) is calculated, combined with the maximum vertical temperature gradient within two meters above the area, and the weighted sum is calculated with weights 0.7 and 0.3 to obtain the local thermal discomfort index (LTI).

[0164] The area with LTI > 1.5℃ / m is selected as the heat sensitive control area, and the jet flow coverage range of the rotational flow air supply outlet is matched to ensure the control pertinence.

[0165] Step 3, Adjustment of Instruction Generation: Adaptation of Cyclone Flow Air Supply Characteristics:

[0166] Air Supply Angle Adjustment Instruction:

[0167] Establish a spatial mapping relationship table between the cyclone flow air outlet and the heat-sensitive regulation area. The trajectory and coverage of the cyclone jet flow are calibrated through CFD numerical simulation (Realizable k-epsilon turbulence model, covering summer cooling and winter heating modes).

[0168] Determine the adjustment direction according to the vertical position of the heat-sensitive regulation area: the lower 1 / 3 height area, the air supply angle is deflected downward (to enhance the sinking effect of the cyclone flow); the middle area remains horizontal with slight adjustment; the upper 1 / 3 height area, deflects upward (to suppress the accumulation of hot air, and to match the cyclone flow induction characteristics).

[0169] Deflection Angle , The local thermal discomfort index, the maximum deflection is ±30°.

[0170] Air Supply Quantity Distribution Instruction:

[0171] According to the personnel density and area of the heat-sensitive regulation area, proportionally increase the air volume of the corresponding cyclone air outlet: increment = baseline air volume × (personnel density / 0.5 person / cubic meter), maximum increment ≤ baseline air volume 50%.

[0172] The air volume instruction is converted into a fan frequency signal (variable frequency range 25-50Hz), and is issued to the variable frequency fan controller, with a response time ≤30 seconds, matching the small air volume and high efficiency air supply characteristics of the cyclone flow system.

[0173] Step 4, Actuator and Closed-loop Control:

[0174] Variable Frequency Fan Control:

[0175] Centrifugal fans (rated air volume 50,000 m³ / h, rated total pressure 800 Pa) are used to change the speed by adjusting the power supply frequency, with an air volume adjustment range of 50%-100%, matching the energy optimization goal of the cyclone flow induction type air supply.

[0176] Step 5, Self-adaptive Adjustment Mechanism:

[0177] After completing one adjustment, the system continuously monitors at a cycle of 15 seconds, repeating the data collection and execution control process.

[0178] After three consecutive adjustments, the heat-sensitive regulation area LTI is less than the threshold, and the angle adjustment is suspended while only maintaining the air volume fine adjustment; when the personnel density change rate is detected to be >20%, a new round of adjustment is immediately triggered to adapt to the dynamic response requirements of the cyclone flow system in personnel flow scenarios.

[0179] It has to be noted that, in the present document, the terms "first", "second", etc. merely serve to identify a subject or action, without necessarily requiring or implying any such actual relationship or order between such subjects or actions. Moreover, the terms "comprising", "containing", or any other similar term are intended to encompass non-exclusive inclusions, such that a process, method, article, or apparatus that comprises a list of elements does not include those elements solely, but can include other elements not expressly listed, or can include elements inherent in such process, method, article, or apparatus.

[0180] While embodiments of the present application have been shown and described, it is to be understood that various modifications, substitutions, combinations, and variations can be undertaken by those skilled in the art without departing from the spirit and scope of the present application, which is defined by the appended claims and their equivalents.

Claims

1. A method for intelligent adjustment of air supply flow organization of high and large space air conditioning, characterized in that, include: Acquiring three-dimensional thermal environment field data within a high-ceilinged space, acquiring spatial distribution density data of personnel, and acquiring current operating parameters of the air conditioning system, wherein acquiring the three-dimensional thermal environment field data within the high-ceilinged space includes: The air temperature values ​​at each measuring point are collected by a distributed temperature sensor array deployed at different heights within a large space. The distributed temperature sensor array is divided into no less than five heights along the vertical direction, and each layer is evenly arranged in a grid pattern on the horizontal plane. The surface temperature field of the main activity areas of people in the high space is scanned non-contactly by infrared thermal imager to obtain the radiation temperature distribution of walls, seats, floors and human body surfaces. By combining air temperature data and radiation temperature data, a continuous three-dimensional temperature field model is reconstructed using an inverse distance weighted interpolation algorithm. Based on the three-dimensional thermal environment field data and the spatial distribution density data of personnel, a spatial thermal comfort evaluation model is constructed, which includes vertical temperature gradient, local heat load intensity, and thermally sensitive areas of personnel. The acquisition of the spatial distribution density data of personnel includes: A wide-angle visible light camera array installed at the top of the tall space is used to capture real-time video images of the entire venue. Foreground segmentation and human target detection are performed on the video images, and a deep learning-based target detection model is used to identify human contours in the images. Based on the pixel position of the human body outline in the image and the known intrinsic and extrinsic parameters of the camera, the two-dimensional image coordinates are mapped to three-dimensional spatial coordinates through perspective transformation. The number of effective human targets within a unit volume of space is counted, and a voxel map of the spatial density distribution of people in cubic meters is generated. Based on the output of the spatial thermal comfort evaluation model, air supply angle adjustment instructions and air supply volume allocation instructions are generated for each air supply outlet. The construction of the spatial thermal comfort evaluation model, which includes vertical temperature gradient, local heat load intensity, and thermally sensitive areas of personnel, includes: Calculate the temperature difference between any two adjacent height layers to form a temperature gradient vector in the vertical direction; Define densely populated areas; Within the densely populated area, calculate the absolute value of the deviation between the average temperature of the area and the set comfortable temperature; Obtain the local thermal discomfort index; The absolute value of the temperature deviation is weighted and summed with the vertical temperature gradient within two meters above the area, with weighting coefficients of 0.7 and 0.3 respectively, to obtain the local thermal discomfort index. Traverse all densely populated areas and filter out areas where the local thermal discomfort index is greater than a preset threshold, marking them as thermally sensitive control zones; The air supply angle adjustment command and air supply volume distribution command are sent to the corresponding variable frequency fan controller to complete the dynamic adjustment of the air supply airflow organization. The generation of air supply angle adjustment commands and air supply volume distribution commands for each air outlet includes: Establish a spatial mapping table between the air outlet and the heat-sensitive control zone. The mapping table is pre-calibrated based on the physical location of the air outlet, the jet coverage range, and the airflow attenuation characteristics. For each heat-sensitive control zone, query its associated set of air outlets; Based on the vertical position of the heat-sensitive control zone, determine the initial adjustment direction of the air supply angle: If the area is located in the lower 1 / 3 of the space, the air supply angle will be deflected downwards; If it is in the middle, keep it horizontal or make minor adjustments; If it is located at the top, it will deflect upwards to prevent the accumulation of hot air; The specific deflection of the air supply angle is obtained by linear mapping of the local thermal discomfort index; At the same time, based on the population density and area within the heat-sensitive control zone, the air supply volume of the associated air outlets in that area is increased proportionally.

2. The intelligent adjustment method for airflow organization in high-ceilinged spaces according to claim 1, characterized in that, The acquisition of the current operating parameters of the air conditioning system includes: Read the current air volume, static pressure, air supply temperature and fan speed of each air supply branch from the building automation system interface; Read the opening status and feedback signals of each electric regulating valve; Read the linkage status of the fresh air valve and the return air valve.

3. The intelligent adjustment method for airflow organization in high-ceilinged spaces according to claim 2, characterized in that, The spatial mapping table is pre-calibrated through CFD numerical simulation. The simulation conditions include two modes: summer cooling and winter heating. The turbulence model adopts the Realizable k-epsilon model.

4. The intelligent adjustment method for airflow organization in high-ceilinged spaces according to claim 3, characterized in that, The deep learning-based object detection model adopts an improved YOLOv5 architecture, with the backbone network replaced by the lightweight MobileNetV3.

5. The intelligent adjustment method for airflow organization in high-ceilinged spaces according to claim 4, characterized in that, After receiving the air volume distribution command, the variable frequency fan controller changes the fan speed by adjusting the power supply frequency of the fan motor.

6. The intelligent adjustment method for airflow organization in high-ceilinged spaces according to claim 5, characterized in that, After completing a dynamic adjustment of the airflow organization, the system enters a continuous monitoring state, repeatedly executing the entire process of data acquisition, model building, instruction generation and execution control, forming a closed-loop adaptive adjustment mechanism. If the local thermal discomfort index of the heat-sensitive control area is less than the preset threshold after three consecutive adjustments, then the angle adjustment is paused and only the air volume is finely adjusted. When the rate of change of the voxel map of personnel density distribution is detected to be greater than the specified value, a new round of adjustment process is immediately triggered.

Images