Hospital clean area dynamic monitoring and regulation method based on laser radar

Abstract

The application discloses a hospital clean area dynamic monitoring and regulation method based on a laser radar, relates to the technical field of monitoring and analysis, and comprises the following steps: aggregating adjacent potential heat source grid units to generate at least one potential heat source area data; in response to the fact that the continuous time length information exceeds a predetermined time length, controlling the laser radar to increase the scanning frequency for fine scanning in the space range corresponding to the area position information and generating fine point cloud data; determining the actual fluctuation intensity value of the potential heat source area data at the current time; if the actual fluctuation intensity value continuously exceeds the preset fluctuation intensity threshold value, sending a device state query instruction to a building automatic control system, receiving the device state data of the corresponding area position information returned by the building automatic control system; and generating a heat source removal instruction according to the device state data. The application has the effect of improving the dynamic monitoring and regulation efficiency.

Description

Technical Field

[0001] This application relates to the field of monitoring and analysis technology, and in particular to a method for dynamic monitoring and control of clean areas in hospitals based on lidar. Background Technology

[0002] Hospital clean areas (such as operating rooms and sterile wards) have extremely stringent requirements for environmental cleanliness, necessitating the maintenance of specific airflow patterns, temperature and humidity levels, and air cleanliness grades. Currently, the microenvironment control of these areas primarily relies on constant air conditioning and purification systems and regular cleaning and maintenance schedules. The air conditioning system operates continuously according to preset temperature and humidity thresholds and fixed air supply parameters, circulating and filtering indoor air through high-efficiency filters to maintain positive pressure and laminar or turbulent flow patterns; while cleaning is carried out according to a fixed schedule, such as comprehensive cleaning before and after daily surgeries and deep cleaning weekly. This control mode based on fixed rules and timed triggers constitutes the basic technical framework for current hospital clean environment management.

[0003] However, the aforementioned traditional control methods have significant technical limitations. Their monitoring systems rely on sparsely distributed temperature, humidity, and differential pressure sensors, which can only reflect changes in the average values ​​of macroscopic environmental parameters and cannot detect localized microscopic disturbances. For example, when clean areas are unattended at night, many medical devices, although switched to standby mode, still generate continuous heat within their internal components; surgical lights and sterilization equipment left over from daytime surgeries also slowly release residual heat. These heat sources heat the surrounding air, forming upward thermal plumes that disturb the previously stable airflow, potentially causing settled particles to re-suspend or altering the distribution of airborne particles. Because such disturbances are microscopic and do not change the overall temperature and humidity averages, conventional sensors cannot detect them, and therefore the air conditioning system will not respond. By the time personnel enter the next day, the cleanliness may have been affected unnoticed. Furthermore, scheduled cleaning systems cannot cope with sudden or cumulative surface dust accumulation. Static dust on floors and equipment surfaces continues to accumulate during unattended periods, becoming a secondary source of contamination once people move or airflow changes occur. The existing technologies commonly suffer from insufficient sensing capabilities, overly coarse control granularity, and delayed passive response, resulting in numerous technical vulnerabilities in the microenvironment control of clean areas in hospitals.

[0004] To address the aforementioned issues, there is an urgent need in this field for an intelligent monitoring and control method capable of dynamic sensing at the microscale, accurate identification of unnecessary heat sources, and proactive intervention and regulation, in order to fill the gap in existing technologies for monitoring and handling of covert airflow disturbances during nighttime, unoccupied periods. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this application provides a method for dynamic monitoring and control of clean areas in hospitals based on lidar.

[0006] The method for dynamic monitoring and control of clean areas in hospitals based on lidar provided in this application includes the following steps: Acquire raw point cloud data generated by periodic scanning of a clean area by lidar during unmanned nighttime hours; The original point cloud data is divided into three-dimensional grid cells of a predetermined number to generate gridded point cloud data; Based on the gridded point cloud data, determine the fluctuation intensity value and occurrence frequency value of each three-dimensional grid cell within a preset time window; Three-dimensional mesh cells whose fluctuation intensity value is greater than a preset fluctuation intensity threshold and whose occurrence frequency value is greater than a preset occurrence frequency threshold are marked as potential heat source mesh cells; Adjacent potential heat source grid cells are aggregated to generate at least one potential heat source region data, which includes region location information and duration information. In response to the duration information exceeding a predetermined duration, the lidar is controlled to increase its scanning frequency for finer scanning within the spatial range corresponding to the area location information, thereby generating fine point cloud data. Based on the refined point cloud data, determine the actual fluctuation intensity value of the potential heat source area data at the current moment; If the actual fluctuation intensity value continues to be greater than the preset fluctuation intensity threshold, a device status query instruction is sent to the building automation system, and the device status data corresponding to the area location information returned by the building automation system is received. Based on the device status data, a heat source removal command is generated. The heat source removal command is used to control the environmental control equipment in the clean area to perform a preset heat source removal operation.

[0007] Preferably, based on the meshed point cloud data, the fluctuation intensity value and occurrence frequency value of each three-dimensional mesh unit within a preset time window are determined, specifically including: Obtain the gridded point cloud data for N consecutive frames within the preset time window, where N is a preset positive integer; For each of the three-dimensional grid cells, the position variance is calculated based on the three-dimensional coordinates of the gridded cloud data at the midpoint of each frame, and the average value of the position variance of N frames is determined as the fluctuation intensity value. For each of the three-dimensional mesh units, the percentage of frames in the N frames where the number of meshed point clouds is greater than a preset threshold is counted, and the percentage of frames is determined as the frequency of occurrence.

[0008] Preferably, adjacent potential heat source grid cells are aggregated to generate at least one potential heat source region data, specifically including: The potential heat source grid cells that are spatially adjacent are clustered using a connected component analysis algorithm to generate at least one connected region; Record the three-dimensional coordinates of the boundary of each connected region to generate region location information; Record the duration of each connected region from the moment it is first marked as a potential heat source grid cell to the current moment to generate duration information.

[0009] Preferably, in response to the duration information exceeding a predetermined duration, the lidar is controlled to increase its scanning frequency for a more refined scan of the spatial range corresponding to the area location information, generating refined point cloud data, specifically including: The duration information is compared with a preset confirmation duration threshold; If the duration information is greater than the preset confirmation duration threshold, a scan frequency adjustment command is generated; The scanning frequency adjustment command and the area location information are sent to the lidar, controlling the lidar to focus and scan the spatial range corresponding to the area location information at a frequency higher than the periodic scanning frequency, thereby generating refined point cloud data.

[0010] Preferably, based on the refined point cloud data, the actual fluctuation intensity value of the potential heat source region data at the current moment is determined, specifically including: Acquire M consecutive frames of refined point cloud data, where M is a preset positive integer; For each frame, the variance of the three-dimensional coordinates of all points in the refined point cloud data is calculated as the regional fluctuation value for each frame. The average value of the regional fluctuation values ​​in the M frames is determined as the actual fluctuation intensity value.

[0011] Preferably, if the actual fluctuation intensity value continuously exceeds the preset fluctuation intensity threshold, a device status query command is sent to the building automation system, and device status data corresponding to the area location information returned by the building automation system is received, specifically including: At multiple consecutive monitoring moments, the actual fluctuation intensity value is compared with the preset fluctuation intensity threshold. If the actual fluctuation intensity value at all monitoring times within the preset confirmation time period is greater than the preset fluctuation intensity threshold, then a device status query instruction carrying the regional location information is generated. Send the device status query command to the building automation system; The system receives the device status data queried and returned by the building automation system based on the area location information. The device status data includes the operating mode information of the device at the corresponding location.

[0012] Preferably, a heat source removal command is generated based on the device status data, specifically including: Analyze the device status data to obtain the operating mode information of all medical devices within the area location information; If there is a medical device in standby mode, a first heat source removal command is generated. The first heat source removal command is used to send a command to the building automation system to switch the corresponding medical device to deep sleep or power-off mode. If no medical device is in standby mode, a second heat source removal command is generated. This second heat source removal command is used to send an instruction to the environmental control device to adjust the air supply parameters, so as to guide the heat plume within the spatial range corresponding to the potential heat source area data to the return air vent.

[0013] Preferably, the generation of a second heat source removal command specifically includes: Obtain the center coordinates of the potential heat source area and the location coordinates of the return air vent within the clean area; Calculate the airflow guidance direction parameters based on the center coordinates and the position coordinates of the return air vent; Based on the airflow guidance direction parameters, an air supply adjustment command is generated, which includes the identifier of the target air supply outlet, the baffle angle adjustment value, and the wind speed adjustment value.

[0014] Preferably, it also includes a step to verify the heat source removal effect: After generating and sending the first heat source removal command or the second heat source removal command, the refined point cloud data is continuously acquired; Based on the refined point cloud data, determine the actual fluctuation intensity value after clearing; The actual fluctuation intensity value after clearing is compared with the preset fluctuation intensity threshold. If the actual fluctuation intensity value after the clearing is less than the preset fluctuation intensity threshold, a successful regulation record is generated and the periodic scan is resumed; If the actual fluctuation intensity value after the clearing is still greater than or equal to the preset fluctuation intensity threshold, then the heat source clearing command is regenerated and an alarm message is generated.

[0015] Preferably, before generating the first heat source removal command, a standby device priority evaluation step is also included: After obtaining the device status data corresponding to the area location information returned by the building automation system, the device identifier and estimated wake-up time data of the medical device in standby state are obtained by parsing. The system queries a preset device management database based on the device identifier to obtain the clinical use priority coefficient and deep sleep energy consumption parameters of each standby device. The clinical use priority coefficient is used to characterize the probability that the device will be urgently activated the next morning. For each standby device, the remaining standby time is calculated based on the difference between the estimated wake-up time data and the current system time, and the estimated energy saving data that could be saved if the power is cut off immediately is calculated in combination with the deep sleep energy consumption parameters. The clinical use priority coefficient, the estimated energy saving data, and the duration information are input into a preset multi-objective decision model to obtain the clearance priority score for each standby device. The removal priority score is compared with a preset removal threshold. If the removal priority score of a standby device is greater than the preset removal threshold, a first heat source removal command is generated for the standby device. If the removal priority score of a standby device is not greater than the preset removal threshold, a temporary airflow guidance command is generated. The temporary airflow guidance command is used to control the environmental control device to guide the hot plume to the return air vent in a minimum energy consumption mode, and to re-evaluate after a preset delay time.

[0016] Preferably, after generating the air supply adjustment command, an adaptive optimization step for the airflow guidance effect is also included: After executing the air supply adjustment command, the refined point cloud data is acquired, and the actual fluctuation intensity value sequence after guidance is determined based on the refined point cloud data. Obtain the actual operating parameters of the current air outlet, including the actual guide vane angle and the actual wind speed; The attenuation rate is calculated based on the actual fluctuation intensity value sequence. If the attenuation rate does not reach the preset target attenuation threshold within the preset guidance effect evaluation period, then an adaptive parameter optimization loop is initiated. Adjust the step size according to the preset parameters, gradually adjust the angle of the guide vane and the actual wind speed, and reacquire the refined point cloud data and calculate the new attenuation rate after each adjustment; Repeat the above adjustment process until the preset maximum number of iterations is reached or the attenuation rate reaches the preset target attenuation threshold. Record the final determined actual operating parameters of the air outlet as the optimal guiding parameters of the potential heat source area and store them in the historical database.

[0017] In summary, this application includes at least one of the following beneficial technical effects: 1. This application provides a method for dynamic monitoring and control of clean areas in hospitals based on lidar. By processing lidar point cloud data during nighttime unattended periods into a grid and calculating the fluctuation intensity and frequency of each three-dimensional grid cell within a time window, it can accurately capture weak airflow disturbances caused by equipment standby heating or residual heat release. This kind of dynamic monitoring at the microscale effectively solves the technical blind spot of not being able to detect changes in airflow organization caused by local heat sources, and can detect potential cleanliness risks even when the overall temperature and humidity do not change significantly. 2. Potential heat source regions are formed by aggregating adjacent potential heat source grid units and recording their duration. Only after the duration exceeds a predetermined time is the lidar controlled to increase the scanning frequency for fine-grained scanning and verification. When the fine-grained scanning confirms that the actual fluctuation intensity continues to exceed the standard, subsequent control actions are triggered. The confirmation logic, which proceeds from coarse to fine and layer by layer, eliminates instantaneous interference or scanning noise, ensuring that only truly continuous abnormal heat sources are monitored, thereby guaranteeing the reliability of the monitoring system and the accuracy of the control actions. 3. By comprehensively considering multiple factors such as the clinical use priority of equipment, estimated energy saving, and the duration of heat source, and inputting these factors into a multi-objective decision-making model for calculation, the most reasonable decision can be made. For non-urgent and energy-intensive equipment, power is cut off first to remove the heat source and save energy; for equipment with high clinical backup priority, power is not cut off temporarily, and temporary airflow guidance measures are adopted. The above mechanism perfectly balances the relationship between ensuring medical emergency response and reducing energy consumption and maintaining a clean environment, avoiding the risk of affecting medical treatment due to blind power outages. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart of a method for dynamic monitoring and control of clean areas in hospitals based on lidar, according to an embodiment of this application. Detailed Implementation

[0020] The following description, in conjunction with the implementation of this invention, is merely an example and illustration of the concept of this invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the inventive concept or exceed the scope defined in these claims, all of which should fall within the protection scope of this invention.

[0021] After introducing the basic concept of the present invention, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Example

[0022] This application discloses a method for dynamic monitoring and control of clean areas in hospitals based on lidar.

[0023] Reference Figure 1 A method for dynamic monitoring and control of clean areas in hospitals based on lidar includes the following steps: Acquire raw point cloud data generated by periodic scanning of a clean area by lidar during unmanned nighttime hours; The original point cloud data is divided into three-dimensional grid cells of a predetermined number to generate gridded point cloud data; Based on the gridded point cloud data, determine the fluctuation intensity value and occurrence frequency value of each three-dimensional grid cell within a preset time window; Three-dimensional mesh cells whose fluctuation intensity value is greater than a preset fluctuation intensity threshold and whose occurrence frequency value is greater than a preset occurrence frequency threshold are marked as potential heat source mesh cells; Adjacent potential heat source grid cells are aggregated to generate at least one potential heat source region data, which includes region location information and duration information. In response to the duration information exceeding a predetermined duration, the lidar is controlled to increase its scanning frequency for finer scanning within the spatial range corresponding to the area location information, thereby generating fine point cloud data. Based on the refined point cloud data, determine the actual fluctuation intensity value of the potential heat source area data at the current moment; If the actual fluctuation intensity value continues to be greater than the preset fluctuation intensity threshold, a device status query instruction is sent to the building automation system, and the device status data corresponding to the area location information returned by the building automation system is received. Based on the device status data, a heat source removal command is generated. The heat source removal command is used to control the environmental control equipment in the clean area to perform a preset heat source removal operation.

[0024] Furthermore, based on the gridded point cloud data, the fluctuation intensity value and occurrence frequency value of each three-dimensional grid cell within a preset time window are determined, specifically including: Obtain the gridded point cloud data for N consecutive frames within the preset time window, where N is a preset positive integer; For each of the three-dimensional grid cells, the position variance is calculated based on the three-dimensional coordinates of the gridded cloud data at the midpoint of each frame, and the average value of the position variance of N frames is determined as the fluctuation intensity value. For each of the three-dimensional mesh units, the percentage of frames in the N frames where the number of meshed point clouds is greater than a preset threshold is counted, and the percentage of frames is determined as the frequency of occurrence.

[0025] In a specific embodiment, after completing the spatial gridding of the original point cloud data, the wave feature extraction stage is entered, which aims to quantify two core feature indicators from the spatiotemporal distribution of the point cloud in each three-dimensional grid cell: wave intensity value and occurrence frequency value, to characterize the activity level of the airflow in the region and the stability of continuous particle filling. First, a sliding time window is constructed, containing N consecutive frames of gridded point cloud data, where N is a preset positive integer. Considering the scanning frequency of the lidar and the typical time scale of airflow disturbance, in this embodiment, N is preferably set to 10, corresponding to a 5-second monitoring window (if the scanning frequency is 2Hz). For each 3D grid cell (denoted as grid j), the 3D coordinates of all valid point clouds within that frame are extracted frame by frame. Let the point cloud coordinates within grid j in the i-th frame be set. Where m is the number of point clouds in grid j within the frame. To quantify the dispersion of the point clouds within the i-th frame, the positional variance of the point clouds is calculated. Specifically, the centroid coordinates of all point clouds within the frame are first calculated. , , Then, the squared Euclidean distance from each point cloud to the centroid is calculated and the mean value is taken to obtain the position variance of the grid j in that frame. The unit is square meters (m²). This variance value reflects the degree of particle dispersion within the grid at that instant. The larger the variance, the more intense the particle movement, suggesting the presence of airflow disturbance. When the number of valid point clouds m in grid j within a frame is lower than a preset threshold, it is determined that there is no valid data for that grid in that frame, and the variance calculation for that frame is skipped and marked as an invalid frame.

[0026] After completing the calculations for all N frames within the time window, for each grid j, the position variances corresponding to all valid frames (i.e., frames with a point cloud count ≥ a preset threshold) are selected. Calculate their arithmetic mean, which is used as the fluctuation intensity value of grid j within the current time window. ,in Number of valid frames. Fluctuation intensity value. The unit is square meters, and its value directly represents the turbulence intensity of the airflow in that area.

[0027] At the same time, the frequency of valid point clouds appearing in grid j across N frames is counted. Specifically, the number of frames in N frames where the number of point clouds within that grid exceeds a preset threshold is counted. The ratio of this ratio to the total number of frames N is used to calculate the frequency of occurrence of grid j. Frequency of occurrence This is a dimensionless value, ranging from [0,1]. Its physical significance lies in distinguishing between continuous airflow phenomena and occasional noise. For example, if particles continuously pass through a grid due to a thermal plume, their occurrence frequency will be close to 1; if they are just random particles that drift by briefly, their occurrence frequency will be lower.

[0028] The above calculation process ensures that both the fluctuation intensity value and the occurrence frequency value originate from the same set of gridded point cloud data and are interrelated: the fluctuation intensity value is calculated based on the spatial scattering within the effective frame, while the occurrence frequency value determines which frames are included in the fluctuation intensity calculation. Together, they constitute the quantitative basis for identifying potential heat sources. For example, within a preset time window (5 seconds, N=10), if the fluctuation intensity value of a certain grid remains consistently high and the occurrence frequency value is also high, it means that there is a continuous and active airflow disturbance in that area, which is very likely caused by a thermal plume, and thus it is marked as a potential heat source grid cell. By introducing a sliding window mechanism, the above indicators can be recalculated with each updated frame of data, achieving dynamic tracking.

[0029] Furthermore, adjacent potential heat source grid cells are aggregated to generate at least one potential heat source region data, specifically including: The potential heat source grid cells that are spatially adjacent are clustered using a connected component analysis algorithm to generate at least one connected region; Record the three-dimensional coordinates of the boundary of each connected region to generate region location information; Record the duration of each connected region from the moment it is first marked as a potential heat source grid cell to the current moment to generate duration information.

[0030] In one specific embodiment, after extracting the fluctuation features and thresholding the three-dimensional mesh elements, the system obtains a set of discrete meshes labeled as potential heat source mesh elements, denoted as... Each grid cell is defined by the coordinates of its center point. Unique identifiers, with coordinates in meters. These discrete grids may correspond to multiple independent heat source regions, such as standby equipment in one corner of an operating room and residual heat from a surgical lamp on the other side. To aggregate spatially adjacent grids into physically meaningful independent heat source regions, the system employs a three-dimensional connected component analysis algorithm for clustering.

[0031] Specifically, firstly, a three-dimensional binary matrix is ​​constructed, whose dimensions correspond to the grid division of the entire clean area, and the elements belonging to... The first grid cell is assigned a value of 1, and the rest are assigned a value of 0. Then, the 26-neighborhood connectivity criterion (i.e., in 3D space, two grid cells are considered adjacent if they share a vertex, an edge, or a face) is used to label the connected components. The algorithm flow is as follows: Initialize a labeling matrix of the same size as the binary matrix. All elements are initially set to 0, and the current connected component number is set. Traverse all grid cells. When encountering a grid cell with a value of 1 that has not yet been labeled, use it as a seed point and assign it a connected component number. It recursively explores all 26 neighboring grid cells with a value of 1 using a breadth-first search algorithm, and marks all explored grid cells as having the same connected component number. After completing the exploration of a connected component, Increment by 1 and continue iterating until all grid cells with a value of 1 have been marked. The output of this algorithm is a set of connected components. , where each connected component It contains a set of spatially adjacent potential heat source grid cells.

[0032] For each connected component Extracting location information from the execution area. Specifically, traversing... The center coordinates of all grid cells The minimum and maximum values ​​along each coordinate axis are taken to form the spatial bounding box of the connected region. The region's location information is represented as... All units are meters. This boundary box is used to define the spatial range during subsequent fine-tuning scans, ensuring that the LiDAR performs high-frequency focused scanning only on this area, rather than scanning the entire area, thereby reducing the system load.

[0033] Meanwhile, the system records the duration of each connected component. To this end, the system maintains a first-occurrence timestamp table, recording the moment when each potential heat source grid cell is first marked as a potential heat source. Let the grid cell... The first labeling time is The unit is seconds. For connected components Its duration information Defined as all grid cells within the connected domain from the time they were first labeled to the current time. The maximum value in the cumulative time is calculated using the following formula: The unit is seconds. The necessity of this design lies in the fact that the heat source region may expand or shrink over time. Using the maximum duration ensures that the system remains sensitive to the earliest persistent heat source, avoiding time resets due to new grid cells added at the region's edge, thus accurately reflecting the stable existence time of the heat source. For example, if a standby device starts generating heat at 22:00, the grid above it is marked at 22:05, and subsequently, at 22:10, the surrounding grid cells are also marked due to the spread of the heat plume. In this case, the duration of the entire connected domain should be calculated based on the earliest marked grid cell (22:05), i.e. It should be 22:10, not 22:10, so as to accurately reflect the actual duration of the heat source.

[0034] To ensure the accuracy and robustness of duration calculations, the system introduces a forgetting mechanism: if a grid cell is not marked as a potential heat source within multiple consecutive time windows (e.g., three consecutive scans, corresponding to 15 seconds), the record for that grid cell is deleted from the first appearance timestamp table, indicating that the heat source in that area has disappeared. This mechanism avoids the accumulation of errors caused by brief disturbances or noise.

[0035] Furthermore, in response to the duration information exceeding a predetermined duration, for the spatial range corresponding to the area location information, the lidar is controlled to increase its scanning frequency for refined scanning, generating refined point cloud data, specifically including: The duration information is compared with a preset confirmation duration threshold; If the duration information is greater than the preset confirmation duration threshold, a scan frequency adjustment command is generated; The scanning frequency adjustment command and the area location information are sent to the lidar, controlling the lidar to focus and scan the spatial range corresponding to the area location information at a frequency higher than the periodic scanning frequency, thereby generating refined point cloud data.

[0036] In one specific embodiment, after obtaining potential heat source region data through connected component analysis, the system acquires the duration information of each region. and regional location information At this point, the system needs to determine whether the heat source is eligible for fine-grained scanning, i.e., whether the heat source has persisted for a sufficiently long time to eliminate interference from accidental disturbances or transient noise. To this end, a confirmation duration threshold is introduced. As a criterion for judgment.

[0037] Confirmation time threshold The settings are based on the physical time constant required for a thermal plume to develop into a stable, detectable airflow, and the sampling interval of the lidar's periodic scans. Experiments have verified that thermal plumes generated by residual heat from equipment in hospital clean areas typically form a stable updraft after 30 to 60 seconds of continuous heating, and the resulting point cloud fluctuations begin to exhibit regular characteristics. To avoid misjudging occasional disturbances caused by brief personnel entry / exit or instantaneous equipment startup, and to ensure timely response to real heat sources, this embodiment confirms a duration threshold. This parameter is adjustable, ranging from 30 to 300 seconds, with 60 seconds being the recommended optimal value. This threshold can be calibrated on-site during system commissioning based on the airflow exchange rate and lidar scanning frequency of the specific clean area: if the ambient airflow is stable and the scanning frequency is high (e.g., 2Hz), the threshold can be appropriately lowered (e.g., 45 seconds) to improve response sensitivity; if there are many interference sources in the environment, the threshold can be increased (e.g., 120 seconds) to enhance robustness.

[0038] In practice, for each potential heat source region its duration information Compared with the preset confirmation time threshold The comparison is performed by the logic comparison unit of the central controller, outputting a Boolean result. : in, and The unit for all measurements is seconds (s), ensuring consistent dimensions and the validity of the comparison operation. If This indicates that the heat source has not yet reached the required confirmation time. The system temporarily stores it in the observation list and continues to monitor its subsequent development through low-frequency periodic scans, without triggering any actions. If the system determines that there is a persistent heat source in the area that requires close attention, it will immediately enter the fine-grained scanning process.

[0039] When the determination result is true, the system generates a scan frequency adjustment command. This command's data structure contains two core fields: target region identifier and target scan frequency. The target region identifier is the region location information generated in the preceding steps. This is used to specify the focused scanning range of the lidar. Target scanning frequency. The setting is based on the following principles: ensuring the capture of dynamic details of the thermal plume while avoiding data redundancy and excessive equipment wear due to excessively high frequencies. In this embodiment, the base frequency for periodic scanning... Set to 0.2Hz (i.e., one frame every 5 seconds) for continuous monitoring across the entire area. Target frequency for fine-tuning scanning. This is increased to 5 to 10 times the base frequency, with the specific value determined based on the time scale of the thermal plume fluctuations. In this embodiment, the target scanning frequency... The preferred setting is 2Hz (i.e., 2 frames per second, corresponding to one frame every 0.5 seconds), which can completely cover the fluctuation frequency band of the thermal plume. The formula for calculating this frequency value is: in, In this embodiment, the frequency is increased by a factor of [number] The value is 10 (increased from 0.2Hz to 2Hz). The value can be adjusted from 5 to 20 according to actual needs: for more refined time-series analysis (such as for airflow velocity inversion), it can be increased to 15 or 20; if only the presence of heat sources needs to be confirmed, it can be reduced to 5 to save resources. The algorithm for generating the scan frequency adjustment command can be expressed as: in, The preset duration of the fine-tuning scan is set to 300 seconds (5 minutes) in this embodiment to ensure sufficient time to collect multiple sets of data for subsequent confirmation of fluctuation intensity and evaluation of control effects. If the heat source is not eliminated after 300 seconds, the system can decide whether to extend the scan or repeat the process based on the results of subsequent steps.

[0040] Subsequently, the system sends the generated scanning frequency adjustment command to the LiDAR controller via the communication interface. Upon receiving the command, the LiDAR performs the following operations: First, based on the area location information in the command... Calculate the corresponding scanning angle range. Assuming the lidar uses a rotating scanning mechanism, its horizontal rotation angle range is... and vertical pitch angle range The coordinates can be calculated from the spatial bounding box coordinates and the installation position of the lidar using a geometric transformation formula. Specifically, let the lidar be located at coordinates... For each of the eight vertices of the bounding box, the azimuth and elevation angles relative to the lidar are calculated, and the minimum and maximum values ​​are taken as the scanning angle range. Next, the lidar switches its scanning mode from full-area 360° scanning to area-focused scanning only within the aforementioned angle range, and simultaneously changes the scanning frequency from the base frequency... Switch to target frequency .

[0041] Through the above mechanism, a dynamic switch from coarse-grained monitoring of the entire area to fine-grained focusing on key areas is achieved. The technical advantages of the above adaptive scanning strategy are: on the one hand, it avoids data redundancy, increased computational load, and reduced lifespan of the lidar caused by continuous high-frequency scanning; on the other hand, it ensures that sufficient spatiotemporal resolution data support can be obtained at critical moments when fine analysis is required, laying the foundation for subsequent precise control.

[0042] Furthermore, based on the refined point cloud data, the actual fluctuation intensity value of the potential heat source region data at the current moment is determined, specifically including: Acquire M consecutive frames of refined point cloud data, where M is a preset positive integer; For each frame, the variance of the three-dimensional coordinates of all points in the refined point cloud data is calculated as the regional fluctuation value for each frame. The average value of the regional fluctuation values ​​in the M frames is determined as the actual fluctuation intensity value.

[0043] In one specific embodiment, after the lidar performs a detailed scan of the target area and continuously outputs high-frame-rate point cloud data, the actual fluctuation intensity value is calculated. This step aims to extract core indicators that can quantify the intensity of thermal plume turbulence from multiple consecutive frames of detailed point cloud data, providing a quantitative basis for subsequent comparison with a preset fluctuation intensity threshold.

[0044] Specifically, the system first sets a sliding time window containing M consecutive frames of refined point cloud data. The value of M needs to balance the real-time performance of the calculation with statistical stability: if M is too small, it is easily affected by single-frame noise; if M is too large, the response will be lagging. Considering that the refined scanning frequency is preferably 2Hz, and the typical fluctuation period of the thermal plume is about 0.5 seconds to 2 seconds, in this embodiment, M is preferably set to 10, corresponding to a 5-second time window, which can effectively smooth short-term fluctuations while maintaining sensitivity to dynamic changes.

[0045] For the current time t, acquire the refined point cloud data of the preceding M consecutive frames. For each frame i, let the frame contain... There are 1 valid point cloud, and the 3D coordinates of each point cloud are 1. The unit is meters. First, calculate the centroid coordinates of all point clouds in this frame. ): Subsequently, the squared Euclidean distance from each point cloud in the frame to the centroid is calculated, and the sums and averages are taken to obtain the regional fluctuation value of the frame. : The unit is square meters, and its value directly quantifies the dispersion of the point cloud within that frame: when the thermal plume is active, particle motion intensifies, and the point cloud dispersion range expands. Increase; conversely, decrease.

[0046] After calculating the fluctuation values ​​for all valid frames in M ​​frames, the system counts the number of valid frames, denoted as . .like If the number of valid frames is less than the preset minimum threshold (set to M / 2, or 5 frames, in this embodiment), the data quality is considered insufficient, and the system discards the data in that window and waits for the next frame to recalculate. If the value is ≥5, the system calculates the arithmetic mean of the regional fluctuation values ​​of these valid frames, which is taken as the actual fluctuation intensity value F(t) at the current time t: F(t) is expressed in square meters and reflects the average airflow turbulence intensity in the target area within the current time window. This value is compared with a preset fluctuation intensity threshold in subsequent steps to determine whether the heat source persists. Through the above variance averaging calculation based on a sliding window, the airflow disturbance intensity in the target area can be quantified in real time and stably, providing reliable data support for heat source confirmation.

[0047] Furthermore, if the actual fluctuation intensity value continues to exceed the preset fluctuation intensity threshold, a device status query command is sent to the building automation system, and device status data corresponding to the area location information returned by the building automation system is received, specifically including: At multiple consecutive monitoring moments, the actual fluctuation intensity value is compared with the preset fluctuation intensity threshold. If the actual fluctuation intensity value at all monitoring times within the preset confirmation time period is greater than the preset fluctuation intensity threshold, then a device status query instruction carrying the regional location information is generated. Send the device status query command to the building automation system; The system receives the device status data queried and returned by the building automation system based on the area location information. The device status data includes the operating mode information of the device at the corresponding location.

[0048] Furthermore, based on the device status data, a heat source removal command is generated, specifically including: Analyze the device status data to obtain the operating mode information of all medical devices within the area location information; If there is a medical device in standby mode, a first heat source removal command is generated. The first heat source removal command is used to send a command to the building automation system to switch the corresponding medical device to deep sleep or power-off mode. If no medical device is in standby mode, a second heat source removal command is generated. This second heat source removal command is used to send an instruction to the environmental control device to adjust the air supply parameters, so as to guide the heat plume within the spatial range corresponding to the potential heat source area data to the return air vent.

[0049] In one specific embodiment, after obtaining the device status data of the target area through device status query, it is necessary to generate corresponding heat source removal instructions based on the analysis results to achieve precise handling of hidden heat sources. The core of this step lies in distinguishing between two different types of heat sources: residual heat from standby devices and residual heat from objects, and adopting differentiated control strategies such as power-off elimination or airflow guidance for each.

[0050] Specifically, upon receiving device status data from the building automation system, the system first parses the data. The device status data includes a list of devices, with each device element containing a device identifier, operating mode information, and installation coordinates. The system iterates through the device list, extracts the operating mode information for each device, and compares it with a preset operating mode encoding table to determine if any device is in "standby" mode.

[0051] If at least one device is found to be in standby mode during the iteration, the system determines that the current heat source is residual heat from a standby device and generates a first heat source removal command. This command includes a command type field (set to "power_control"), a target device list (identifiers of all standby devices), and an operation type. The operation type is determined based on a preset energy-saving strategy: for devices expected to be unused for an extended period, setting it to "power_off" indicates a complete power cut-off; for devices requiring network connectivity or rapid startup, setting it to "deep_sleep" indicates switching to deep sleep mode. After the first heat source removal command is generated, the system sends it to the building automation system via a communication interface. The building automation system then sends control signals to the power management modules of the corresponding devices to execute the power cut-off or deep sleep operation.

[0052] If no standby devices are found in the device list, meaning all devices are running, powered off, or in deep sleep mode, the system determines the current heat source to be residual heat from an object and generates a second heat source removal command. This command forces the heat plume towards the return air vent via airflow guidance. When generating the second heat source removal command, the system first calculates the center coordinates of the potential heat source area data, obtained by averaging the boundary ranges in the area location information. Then, the system reads the location coordinates of all return air vents within the clean area from the environmental configuration database and calculates the direction vector from the heat source center to the nearest return air vent. Simultaneously, the system obtains the coordinates of the supply air vents above the target area and calculates the horizontal deflection angle and vertical pitch angle of the guide vanes required to guide the airflow from the supply air vents to the heat source center. Furthermore, the system calculates the wind speed adjustment increment based on the ratio of the current actual fluctuation intensity value to a preset fluctuation intensity threshold, and adds this increment to the base supply air speed as the guiding wind speed. The system encapsulates the calculated deflector angle and wind speed into an air supply adjustment command, which includes the command type, target air outlet identifier, deflector horizontal angle, deflector vertical angle, and wind speed value, and sends it to the environmental control equipment to perform airflow guidance operation.

[0053] Through the aforementioned differentiated instruction generation mechanism, the system achieves precise handling of residual heat from standby equipment and residual heat from objects: the former eliminates the heat source at its source by cutting off the power, while the latter directs the heat out through airflow guidance, jointly ensuring the stability of airflow organization during the nighttime unattended period in the clean area.

[0054] Furthermore, a second heat source removal command is generated, specifically including: Obtain the center coordinates of the potential heat source area and the location coordinates of the return air vent within the clean area; Calculate the airflow guidance direction parameters based on the center coordinates and the position coordinates of the return air vent; Based on the airflow guidance direction parameters, an air supply adjustment command is generated, which includes the identifier of the target air supply outlet, the baffle angle adjustment value, and the wind speed adjustment value.

[0055] In one specific embodiment, after determining that the current heat source is residual heat from an object, a second heat source removal command needs to be generated. This command forces the hot plume towards the return air vent by adjusting the air supply parameters. The core of this step lies in converting the spatial location of the heat source into executable angle and wind speed parameters for the air supply vent, thereby achieving precise airflow guidance.

[0056] Specifically, the first step is to obtain the center coordinates of the potential heat source region. These center coordinates are derived from the region location information generated in the preceding steps. The calculation formula is as follows: in, These are the three-dimensional coordinates of the center point, in meters (m). The reason for using the geometric center instead of other locations (such as the centroid) is that thermal plumes typically spread upwards from the center of the heat source, and using the geometric center as the airflow guidance target can cover the maximum thermal impact range.

[0057] Simultaneously, the location coordinates of all return air vents within the clean area are retrieved from the environmental configuration database. These return air vent coordinates are pre-calibrated through on-site measurement or import from a BIM model and stored as three-dimensional coordinate points. The unit is meters. The system calculates the Euclidean distance from the heat source center to each return air vent and selects the closest return air vent as the target return air vent. The necessity of choosing the nearest return air outlet is that the closer the distance, the shorter the airflow guidance path, the less likely the hot plume will diffuse during transmission, and the higher the removal efficiency.

[0058] After identifying the target return air vent, the system needs to calculate how to adjust the supply air vent parameters to generate an airflow that directs the hot plume towards that return air vent. The system obtains the supply air vent identifiers and coordinates above the heat source area. The selection principle for the supply air vents is: the laminar flow air supply unit (FFU) closest to and directly above the center of the heat source. The system then iterates through the coordinates of all supply air vents. Calculate the horizontal projection distance and vertical height difference between it and the center of the heat source, prioritizing the option with the smallest horizontal projection distance and a z-coordinate greater than 1. The air outlets ensure that airflow covers the heat source area from top to bottom.

[0059] Let the coordinates of the selected target air outlet be... Calculate the vector from the air outlet to the center of the heat source: All three components of this vector are in meters. Based on this vector, the system calculates the horizontal deflection angle and vertical pitch angle of the deflector. Horizontal deflection angle Defined as the angle between the positive direction of the air outlet (usually the reference direction calibrated during installation, such as due north) and the horizontal projection of the vector, the calculation formula is: The calculation result is in degrees, and the range of values ​​is... In actual use, the mechanical limit of the air outlet needs to be mapped to... Or a specific adjustable range. If and ,but ;like and ,but Or 270°.

[0060] Vertical pitch angle Defined as the angle between the vector and the horizontal plane, the calculation formula is: The calculation result is in degrees, and the range of values ​​is... A positive value indicates upward airflow, while a negative value indicates downward airflow. This is because the heat source is located below the air outlet. , It is a negative value, therefore A negative angle indicates that the deflector needs to be tilted downwards.

[0061] In addition to the angle parameters, the supply air velocity adjustment value also needs to be determined. To avoid excessive airflow disturbing the overall laminar flow pattern of the clean area, the guiding air velocity should be appropriately increased from the base supply air velocity. Base supply air velocity Based on the preset cleanroom class, for an ISO 5 (Class 100) operating room, the laminar flow air supply velocity is typically 0.3 m / s to 0.5 m / s. In this embodiment, we take... =0.3 Wind speed adjustment value The heat source intensity is dynamically determined based on the current actual fluctuation value. The characterization and calculation formula is as follows: in, The preset wind speed gain coefficient is taken in this embodiment. Ensure that the maximum wind speed does not exceed 0.45 m / s (i.e., 1.5 times the base wind speed) to avoid disrupting the laminar flow pattern; The actual fluctuation intensity value at the current moment, in square meters; The preset fluctuation intensity threshold is expressed in square meters. The formula is designed based on the principle that the stronger the heat source (the greater the fluctuation intensity), the greater the required airflow velocity is needed to ensure sufficient momentum to direct the hot plume towards the return air vent. hour, The guiding wind speed is 0.45 m / s; when hour, The guiding wind speed is 0.60 m / s, but it must be limited to ensure it does not exceed the preset maximum value (in this embodiment, the maximum guiding wind speed is limited to 0.6 m / s). Final guiding wind speed for: in, The maximum permissible wind speed is preset, which is 0.6 m / s in this embodiment.

[0062] Through the above coordinate calculation, angle calculation, and dynamic wind speed adjustment, the spatial location of the heat source is accurately converted into executable parameters of the air outlet, realizing the directional airflow guidance and removal of the residual heat of the object.

[0063] Furthermore, it also includes steps to verify the effectiveness of heat source removal: After generating and sending the first heat source removal command or the second heat source removal command, the refined point cloud data is continuously acquired; Based on the refined point cloud data, determine the actual fluctuation intensity value after clearing; The actual fluctuation intensity value after clearing is compared with the preset fluctuation intensity threshold. If the actual fluctuation intensity value after the clearing is less than the preset fluctuation intensity threshold, a successful regulation record is generated and the periodic scan is resumed; If the actual fluctuation intensity value after the clearing is still greater than or equal to the preset fluctuation intensity threshold, then the heat source clearing command is regenerated and an alarm message is generated.

[0064] Furthermore, before generating the first heat source removal command, a standby device priority evaluation step is also included: After obtaining the device status data corresponding to the area location information returned by the building automation system, the device identifier and estimated wake-up time data of the medical device in standby state are obtained by parsing. The system queries a preset device management database based on the device identifier to obtain the clinical use priority coefficient and deep sleep energy consumption parameters of each standby device. The clinical use priority coefficient is used to characterize the probability that the device will be urgently activated the next morning. For each standby device, the remaining standby time is calculated based on the difference between the estimated wake-up time data and the current system time, and the estimated energy saving data that could be saved if the power is cut off immediately is calculated in combination with the deep sleep energy consumption parameters. The clinical use priority coefficient, the estimated energy saving data, and the duration information are input into a preset multi-objective decision model to obtain the clearance priority score for each standby device. The removal priority score is compared with a preset removal threshold. If the removal priority score of a standby device is greater than the preset removal threshold, a first heat source removal command is generated for the standby device. If the removal priority score of a standby device is not greater than the preset removal threshold, a temporary airflow guidance command is generated. The temporary airflow guidance command is used to control the environmental control device to guide the hot plume to the return air vent in a minimum energy consumption mode, and to re-evaluate after a preset delay time.

[0065] In one specific embodiment, after obtaining the device status data of the target area through device status query and parsing out the existence of standby devices, a power-off command is not directly generated. Instead, the priority of clearing these standby devices needs to be further evaluated. This is because during the nighttime off-peak hours, some devices may be urgently activated in the early morning (such as ventilators and monitors in the ICU), and abruptly cutting off power could affect the normal operation of clinical work the next day. At the same time, different devices have different energy consumption levels and heat source contributions. To address this, the system introduces a standby device priority evaluation mechanism, which comprehensively weighs clinical needs, energy-saving benefits, and heat source severity through a multi-objective decision model to generate differentiated clearing strategies.

[0066] Specifically, the system first parses the device status data returned by the building automation system to obtain a list of device identifiers for medical devices in standby mode. And the estimated wake-up time for each device. The estimated wake-up time data comes from the hospital's scheduling system or equipment usage history. For example, if an anesthesia machine in a certain operating room is usually first activated at 6:30 AM every day, then the estimated wake-up time for that device is set to 6:30 AM. This data is stored in Unix timestamp format in seconds for easy subsequent calculations.

[0067] For each standby device The system queries the pre-set equipment management database based on the equipment identifier. The equipment management database pre-loads various attributes of all medical equipment in the hospital, including: Clinical use priority coefficient This is a dimensionless numerical value, ranging from 1 to 10. A higher value indicates a greater probability that the equipment will be urgently activated the following morning. This coefficient is set by the hospital's equipment management department based on the type of equipment and its clinical importance. For example: 9-10 for life support equipment such as ICU ventilators and monitors; 7-8 for operating room anesthesia machines and high-frequency electrosurgical units; 5-6 for infusion pumps and nutrition pumps in general wards; and 3-4 for auxiliary equipment such as air sterilizers and humidifiers. The logic behind setting this priority coefficient is to ensure that critical equipment will not be unexpectedly powered off due to energy conservation efforts.

[0068] Deep sleep energy consumption parameters The average power consumption of the device in standby mode, measured in watts (W), is obtained from the device nameplate or through actual measurement. For example, the standby power consumption of a certain model of anesthesia machine is 45W, and the standby power consumption of a monitor is 15W.

[0069] At the same time, obtain the current system time. For each device, calculate the remaining standby time. : The remaining standby time is in seconds, representing the time interval from the current moment until the device is expected to be woken up. If This indicates that the expected wake-up time has passed. This indicates that the device should be in operation but is actually still in standby mode, which may indicate an abnormality. In this case, the priority should be increased.

[0070] Based on the remaining standby time and standby power consumption, the system calculates the estimated energy savings that could be achieved if the power is cut off immediately. : The estimated energy savings are measured in joules (J). This value quantifies the energy-saving benefits of power outages: the longer the remaining standby time and the greater the standby power consumption, the greater the energy savings and the higher the benefit of the power outage.

[0071] After calculating the above three basic indicators, the process moves to the multi-objective decision-making stage, where clinical use priority coefficients are assigned. Estimated energy savings and information on the duration of the heat source (That is, the duration from when the potential heat source area was first marked to the current time, in seconds) A comprehensive evaluation is performed to calculate the clearance priority score for each standby device. The formula for calculating the priority score for removal is as follows: The meanings and dimensions of each parameter are as follows: Clinical use priority coefficient, dimensionless, with a value range of 1-10.

[0072] The minimum and maximum priority coefficients among all standby devices, used for normalization.

[0073] Estimated energy saving, in joules (J).

[0074] The minimum and maximum estimated energy savings among all standby devices.

[0075] : Duration of heat source, in seconds (s), derived from the duration information generated in the preceding steps.

[0076] : The minimum and maximum duration of all heat sources currently awaiting processing.

[0077] The preset weighting coefficients are dimensionless and satisfy... In this embodiment, it is preferably set to... These represent the trade-offs between clinical needs, energy-saving benefits, and the severity of heat sources, respectively. The weighting logic is based on the actual needs of hospital operations: clinical needs are the primary consideration, therefore assigned the highest weight of 0.6; energy-saving benefits are secondary, assigned 0.3; and the severity of heat sources is a secondary factor, assigned 0.1.

[0078] The first term in the formula uses This format is because of its higher clinical priority ( The larger the device (and the more important it is), the less likely it is to be powered off; therefore, its removal priority should be lower. High priority is mapped to low score, which aligns with the logic that higher scores should be cleared more frequently. The second and third items use positive normalization, meaning that the greater the energy saving and the longer the heat source duration, the higher the clearing priority.

[0079] Clear priority score The value ranges from 0 to 1, with higher values ​​indicating that the device should be powered off and cleared.

[0080] The calculated priority score for each device is compared with the preset removal threshold. A comparison is then made. The threshold for clearing power is set based on the tolerance for the risk of accidental power outages; in this embodiment, it is preferably set to 0.6. The logic for selecting this value is: only when the comprehensive score exceeds 0.6 is the benefit of the power outage operation considered to significantly outweigh the risk. The threshold can be adjusted within the range of 0.5 to 0.8: if the hospital has high energy-saving requirements, it can be appropriately reduced to 0.5; if it is extremely sensitive to clinical risks, it can be increased to 0.7 or 0.8.

[0081] After comparison, two scenarios emerge: 1. If there is at least one standby device, its If the condition is met, a first heat source removal command is generated for that device, switching it to deep sleep or power-off mode. If multiple devices meet the condition, then... Instructions are generated sequentially from high to low, with only one device operated at a time, and the fluctuation intensity is reassessed after each operation to avoid excessive intervention.

[0082] 2. If all standby devices If the system temporarily suspends power outage, it will instead generate a temporary airflow guidance command. This command controls the environmental control equipment to guide the thermal plume to the return air vent in a minimum energy consumption mode, while simultaneously delaying the flow for a preset time. The system will then be re-evaluated. In this embodiment, the delay time is set to 30 minutes. The selection logic is that 30 minutes is sufficient for some devices about to be activated to enter the pre-wake-up phase, and may also allow temporary heat sources to dissipate naturally, while avoiding frequent evaluations that increase system load. The delay time can be adjusted within the range of 15 to 60 minutes.

[0083] Through the aforementioned multi-objective decision-making mechanism, the system achieves a dynamic balance between energy-saving needs and clinical safety, avoiding accidental power outages to critical equipment, while prioritizing the removal of high-energy-consuming and long-standby non-critical equipment, reflecting the concept of intelligent and refined control.

[0084] Furthermore, after generating the air supply adjustment command, an adaptive optimization step for the airflow guidance effect is also included: After executing the air supply adjustment command, the refined point cloud data is acquired, and the actual fluctuation intensity value sequence after guidance is determined based on the refined point cloud data. Obtain the actual operating parameters of the current air outlet, including the actual guide vane angle and the actual wind speed; The attenuation rate is calculated based on the actual fluctuation intensity value sequence. If the attenuation rate does not reach the preset target attenuation threshold within the preset guidance effect evaluation period, then an adaptive parameter optimization loop is initiated. Adjust the step size according to the preset parameters, gradually adjust the angle of the guide vane and the actual wind speed, and reacquire the refined point cloud data and calculate the new attenuation rate after each adjustment; Repeat the above adjustment process until the preset maximum number of iterations is reached or the attenuation rate reaches the preset target attenuation threshold. Record the final determined actual operating parameters of the air outlet as the optimal guiding parameters of the potential heat source area and store them in the historical database.

[0085] In one specific embodiment, after executing the air supply adjustment command, the actual airflow guidance effect may deviate from the expectation. For example, due to on-site airflow disturbances, equipment response errors, or model simplification, the thermal plume may not be effectively guided to the return air vent. To address this, an adaptive optimization step for airflow guidance effect is introduced. By monitoring changes in fluctuation intensity in real time, the air supply parameters are dynamically adjusted until the expected purging effect is achieved, forming a closed-loop feedback control.

[0086] Specifically, after executing the air supply adjustment command, fine point cloud data of the target area is acquired at a fine scanning frequency (2Hz), and the actual fluctuation intensity value at each moment is calculated according to the aforementioned method to form a sequence of actual fluctuation intensity values ​​after guidance. The sequence is continuously updated using a sliding window method, with the window length consistent with the aforementioned (M=10 frames, corresponding to 5 seconds), ensuring real-time performance.

[0087] At the same time, the current actual operating parameters are obtained from the air outlet actuator, including the actual horizontal angle of the guide vane. (Unit: degrees), actual vertical angle of the guide vane (in degrees) and actual wind speed (Unit: meters per second). These parameters are read through the actuator's feedback interface and used for basic state recording in closed-loop control.

[0088] The system defines the baseline fluctuation intensity value before guidance. This is the actual fluctuation intensity value calculated last before generating the second heat source removal command, i.e., the fluctuation intensity when the heat source's existence is confirmed but before initiation begins. Based on and the real-time fluctuation intensity value sequence after guidance The system calculates the attenuation rate. : The attenuation rate is expressed as a percentage (%), ranging from 0% to 100%. The effect of airflow guidance on suppressing thermal plumes was quantified: the larger the value, the greater the reduction in wave intensity and the better the guidance effect.

[0089] The system sets a time limit for evaluating the effectiveness of the guidance. This parameter defines the waiting time from the start of command execution to the first effect evaluation, avoiding misjudgments due to airflow response delays. Considering that it takes time for airflow to reach the heat source area from the air outlet and produce an effect, this embodiment... The preferred setting is 60 seconds. This value can be adjusted from 30 to 120 seconds depending on the size of the clean area and the airflow speed: the larger the space and the lower the airflow speed, the longer the evaluation time should be.

[0090] exist At the end, the system calculates the decay rate at that moment. and the preset target attenuation threshold A comparison is made. The target attenuation threshold characterizes the desired removal effect, and in this embodiment, it is preferably set to 80%, that is, the expected fluctuation intensity decreases by more than 80%. The logic for setting this threshold is: when the heat plume is effectively guided to the return air vent, the particle concentration and turbulence intensity in the heat source area should decrease significantly, and an attenuation rate of 80% can be considered as the heat source being basically eliminated. The threshold can be adjusted within the range of 70% to 90% according to cleanliness requirements: for ISO Class 5 (Class 100) operating rooms, it can be increased to 85% or 90%; for ISO Class 7 (Class 10,000) areas, it can be appropriately reduced to 70%.

[0091] like If the guidance is successful, the system records the current air supply parameters as the optimal guidance parameters for the heat source location, stores them in the historical database, exits the optimization process, and resumes periodic scanning and monitoring.

[0092] like If the guidance effect is insufficient, the system will initiate an adaptive parameter optimization loop. This loop uses the coordinate descent method to gradually adjust the angle of the guide vane and the wind speed within a preset parameter space. After each adjustment, the attenuation rate is reassessed until the target is reached or the maximum number of iterations is reached.

[0093] The specific implementation of the parameter adaptive optimization loop is as follows: First, define the parameter adjustment step size. (This refers to the horizontal angle step of the deflector.) and vertical angle step size All angles are set to 5°. This value is chosen because a 5° angle change is sufficient to cause a significant change in airflow direction, while avoiding over-adjustment due to excessively large step sizes. The angle step size can be adjusted from 2° to 10° depending on the accuracy of the air outlet. (Wind speed step size...) Set to 0.1 m / s, this value can be adjusted within the range of 0.05 m / s to 0.2 m / s.

[0094] Secondly, define the parameter adjustment range to ensure that it does not exceed the mechanical limits of the air outlet and the maximum allowable air velocity in the clean area. Assume the adjustable range of the air outlet horizontal angle is... The adjustable range of the vertical angle is: The maximum permissible wind speed is (0.6 m / s in this embodiment). The initial parameters are theoretically calculated values. ,in, The wind speed is used to guide the airflow calculated above.

[0095] The optimization process employs an alternating optimization strategy, with each iteration consisting of two steps: First step, fix the current wind speed Optimize the angle of the air deflector. and Within the scope, with and Generate a candidate angle combination mesh for the step size. For each set of candidate angles... The system generates a temporary air supply adjustment command and sends it to the actuator, waiting for the airflow to stabilize. (In this embodiment, it is set to 10 seconds to ensure that the airflow field is re-established), and then refined point cloud data under this parameter is acquired to calculate the new attenuation. After iterating through all candidate angles, select the one that makes... The largest angle combination as the new and If multiple candidate angles produce similar attenuation rates (difference less than 2%), the angle closest to the theoretical initial value will be selected first. A combination of these elements is used to maintain the stability of the airflow path.

[0096] The second step is to fix the optimized angle. Optimize wind speed. Within the scope, with Candidate wind speeds are generated for the step size, where, The base wind speed is 0.3 m / s. For each candidate wind speed... Similarly, waiting Then calculate the new attenuation rate , choose to Maximum wind speed as new .

[0097] After completing one round of angle-wind speed optimization, the system checks whether the termination conditions are met: Condition 1: Current attenuation rate .

[0098] Condition 2: Reach the preset maximum number of iterations In this embodiment, the iteration count is set to 5. The logic behind this setting is: if the target is not reached after 5 iterations, it indicates that there may be other factors (such as the heat source itself has decayed or the equipment has failed), and further optimization is not meaningful, so manual alarm should be triggered.

[0099] If any termination condition is met, the optimization loop ends, and the finally determined air supply parameters are... The optimal guidance parameters for this heat source location are recorded and stored in the historical database. The historical database stores the location of each successfully guided heat source and its corresponding optimal parameters, which can be directly retrieved when a similar heat source appears again in the future, avoiding repeated optimization.

[0100] If the termination condition is not met, the next iteration will begin with the current parameters.

[0101] Through the aforementioned adaptive optimization mechanism, the system can automatically seek optimization when the initial guidance effect is insufficient, finding the optimal combination of air supply parameters for a specific heat source location to ensure that the thermal plume is effectively removed. This closed-loop feedback design significantly improves the system's robustness and adaptability, avoiding control failures caused by model errors or environmental changes.

[0102] The above content is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention, they should all fall within the protection scope of the present invention.

[0103] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0104] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention.

Claims

1. A method for dynamic monitoring and control of clean areas in hospitals based on lidar, characterized in that, Includes the following steps: Acquire raw point cloud data generated by periodic scanning of a clean area by lidar during unmanned nighttime hours; The original point cloud data is divided into three-dimensional grid cells of a predetermined number to generate gridded point cloud data; Based on the gridded point cloud data, determine the fluctuation intensity value and occurrence frequency value of each three-dimensional grid cell within a preset time window; Three-dimensional mesh cells whose fluctuation intensity value is greater than a preset fluctuation intensity threshold and whose occurrence frequency value is greater than a preset occurrence frequency threshold are marked as potential heat source mesh cells; Adjacent potential heat source grid cells are aggregated to generate at least one potential heat source region data, which includes region location information and duration information. In response to the duration information exceeding a predetermined duration, the lidar is controlled to increase its scanning frequency for finer scanning within the spatial range corresponding to the area location information, thereby generating fine point cloud data. Based on the refined point cloud data, determine the actual fluctuation intensity value of the potential heat source area data at the current moment; If the actual fluctuation intensity value continues to be greater than the preset fluctuation intensity threshold, a device status query instruction is sent to the building automation system, and the device status data corresponding to the area location information returned by the building automation system is received. Based on the device status data, a heat source removal command is generated. The heat source removal command is used to control the environmental control equipment in the clean area to perform a preset heat source removal operation.

2. The method for dynamic monitoring and control of hospital clean areas based on lidar according to claim 1, characterized in that, Based on the meshed point cloud data, the fluctuation intensity value and occurrence frequency value of each three-dimensional mesh unit within a preset time window are determined, specifically including: Obtain the gridded point cloud data for N consecutive frames within the preset time window, where N is a preset positive integer; For each of the three-dimensional grid cells, the position variance is calculated based on the three-dimensional coordinates of the gridded cloud data at the midpoint of each frame, and the average value of the position variance of N frames is determined as the fluctuation intensity value. For each of the three-dimensional mesh units, the percentage of frames in the N frames where the number of meshed point clouds is greater than a preset threshold is counted, and the percentage of frames is determined as the frequency of occurrence.

3. The method for dynamic monitoring and control of hospital clean areas based on lidar according to claim 1, characterized in that, Adjacent potential heat source grid cells are aggregated to generate at least one potential heat source region data, specifically including: The potential heat source grid cells that are spatially adjacent are clustered using a connected component analysis algorithm to generate at least one connected region; Record the three-dimensional coordinates of the boundary of each connected region to generate region location information; Record the duration of each connected region from the moment it is first marked as a potential heat source grid cell to the current moment to generate duration information.

4. The method for dynamic monitoring and control of hospital clean areas based on lidar according to claim 1, characterized in that, In response to the duration information exceeding a predetermined duration, the lidar is controlled to increase its scanning frequency for a more refined scan of the spatial range corresponding to the area location information, generating refined point cloud data, specifically including: The duration information is compared with a preset confirmation duration threshold; If the duration information is greater than the preset confirmation duration threshold, a scan frequency adjustment command is generated; The scanning frequency adjustment command and the area location information are sent to the lidar, controlling the lidar to focus and scan the spatial range corresponding to the area location information at a frequency higher than the periodic scanning frequency, thereby generating refined point cloud data.

5. The method for dynamic monitoring and control of hospital clean areas based on lidar according to claim 1, characterized in that, Based on the refined point cloud data, the actual fluctuation intensity value of the potential heat source region data at the current moment is determined, specifically including: Acquire M consecutive frames of refined point cloud data, where M is a preset positive integer; For each frame, the variance of the three-dimensional coordinates of all points in the refined point cloud data is calculated as the regional fluctuation value for each frame. The average value of the regional fluctuation values ​​in the M frames is determined as the actual fluctuation intensity value.

6. The method for dynamic monitoring and control of hospital clean areas based on lidar according to claim 1, characterized in that, If the actual fluctuation intensity value continues to exceed the preset fluctuation intensity threshold, a device status query command is sent to the building automation system, and the device status data corresponding to the area location information returned by the building automation system is received, specifically including: At multiple consecutive monitoring moments, the actual fluctuation intensity value is compared with the preset fluctuation intensity threshold. If the actual fluctuation intensity value at all monitoring times within the preset confirmation time period is greater than the preset fluctuation intensity threshold, then a device status query instruction carrying the regional location information is generated. Send the device status query command to the building automation system; The system receives the device status data queried and returned by the building automation system based on the area location information. The device status data includes the operating mode information of the device at the corresponding location.

7. The method for dynamic monitoring and control of hospital clean areas based on lidar according to claim 1, characterized in that, Based on the device status data, a heat source removal command is generated, specifically including: Analyze the device status data to obtain the operating mode information of all medical devices within the area location information; If there is a medical device in standby mode, a first heat source removal command is generated. The first heat source removal command is used to send a command to the building automation system to switch the corresponding medical device to deep sleep or power-off mode. If no medical device is in standby mode, a second heat source removal command is generated. This second heat source removal command is used to send an instruction to the environmental control device to adjust the air supply parameters, so as to guide the heat plume within the spatial range corresponding to the potential heat source area data to the return air vent.

8. The method for dynamic monitoring and control of hospital clean areas based on lidar according to claim 7, characterized in that, The command to remove the second heat source is generated, specifically including: Obtain the center coordinates of the potential heat source area and the location coordinates of the return air vent within the clean area; Calculate the airflow guidance direction parameters based on the center coordinates and the position coordinates of the return air vent; Based on the airflow guidance direction parameters, an air supply adjustment command is generated, which includes the identifier of the target air supply outlet, the baffle angle adjustment value, and the wind speed adjustment value.

9. The method for dynamic monitoring and control of hospital clean areas based on lidar according to claim 7, characterized in that, It also includes a step to verify the effectiveness of heat source removal: After generating and sending the first heat source removal command or the second heat source removal command, the refined point cloud data is continuously acquired; Based on the refined point cloud data, determine the actual fluctuation intensity value after clearing; The actual fluctuation intensity value after clearing is compared with the preset fluctuation intensity threshold. If the actual fluctuation intensity value after the clearing is less than the preset fluctuation intensity threshold, a successful regulation record is generated and the periodic scan is resumed; If the actual fluctuation intensity value after the clearing is still greater than or equal to the preset fluctuation intensity threshold, then the heat source clearing command is regenerated and an alarm message is generated.

10. The method for dynamic monitoring and control of hospital clean areas based on lidar according to claim 7, characterized in that, Before generating the first heat source removal command, a standby device priority evaluation step is also included: After obtaining the device status data corresponding to the area location information returned by the building automation system, the device identifier and estimated wake-up time data of the medical device in standby state are obtained by parsing. The system queries a preset device management database based on the device identifier to obtain the clinical use priority coefficient and deep sleep energy consumption parameters of each standby device. The clinical use priority coefficient is used to characterize the probability that the device will be urgently activated the next morning. For each standby device, the remaining standby time is calculated based on the difference between the estimated wake-up time data and the current system time, and the estimated energy saving data that could be saved if the power is cut off immediately is calculated in combination with the deep sleep energy consumption parameters. The clinical use priority coefficient, the estimated energy saving data, and the duration information are input into a preset multi-objective decision model to obtain the clearance priority score for each standby device. The removal priority score is compared with a preset removal threshold. If the removal priority score of a standby device is greater than the preset removal threshold, a first heat source removal command is generated for the standby device. If the removal priority score of a standby device is not greater than the preset removal threshold, a temporary airflow guidance command is generated. The temporary airflow guidance command is used to control the environmental control device to guide the hot plume to the return air vent in a minimum energy consumption mode, and to re-evaluate after a preset delay time.

11. The method for dynamic monitoring and control of hospital clean areas based on lidar according to claim 8, characterized in that, After generating the air supply adjustment command, the process also includes an adaptive optimization step for the airflow guidance effect: After executing the air supply adjustment command, the refined point cloud data is acquired, and the actual fluctuation intensity value sequence after guidance is determined based on the refined point cloud data. Obtain the actual operating parameters of the current air outlet, including the actual guide vane angle and the actual wind speed; The attenuation rate is calculated based on the actual fluctuation intensity value sequence. If the attenuation rate does not reach the preset target attenuation threshold within the preset guidance effect evaluation period, then an adaptive parameter optimization loop is initiated. Adjust the step size according to the preset parameters, gradually adjust the angle of the guide vane and the actual wind speed, and reacquire the refined point cloud data and calculate the new attenuation rate after each adjustment; Repeat the above adjustment process until the preset maximum number of iterations is reached or the attenuation rate reaches the preset target attenuation threshold. Record the final determined actual operating parameters of the air outlet as the optimal guiding parameters of the potential heat source area and store them in the historical database.

Images